Press Releases
Principal Investigator: Prof Mahmoud soliman
Institution Name: University of KwaZulu-Natal
Active Member Count: 37
Allocation Start: 2021-01-04
Allocation End: 2021-07-03
Used Hours: 739987
Project Name: Drug Design, development and modelling
Project Shortname: HEAL0790
Discipline Name: Health Sciences
The research scope of the Molecular Bio-Computation and Drug Design Research Lab covers a wide range of computational and molecular modeling research areas with main focus on biological systems and drug design approaches. Main interest is related to design and study of biologically and therapeutically oriented targets by employing the applications of computational methods to the study of problems of chemical and biochemical reactivity, with particular focus upon the transition state, environmental effects on mechanisms, the origins of catalysis, and the interpretation of kinetic isotope effects. This includes mechanistic pathways and transition states for reactions in enzyme and solutions; design of enzyme inhibitors and exploring the binding and catalytic theme of the designed targets and adopting sophisticated computational approaches to understand protein structures and functions. We show keen interest in diseases of global burden and of high prevalence within the south African populace such as HIV/AIDS, Tuberculosis and Cancer. The computational resources provided by CHPC is employed in performing molecular dynamic calculations and the accompanying post molecular dynamic simulation analysis.
Principal Investigator: Prof David Lokhat
Institution Name: University of KwaZulu-Natal
Active Member Count: 1
Allocation Start: 2021-01-04
Allocation End: 2021-07-03
Used Hours: 133824
Project Name: Carbon dioxide hydrogenation
Project Shortname: CHEM1384
Discipline Name: Chemical Engineering
The project is about the development of heterogenous coagulants for water treatment which needs some background investigations. The research programme is greatly enhanced through the use of HPC. The published work and the on-going ones were faster to achieve with HPC than the local machine. There are many software packages available on CHPC which we are making use of and have fast tracked the research progress.
Principal Investigator: Prof Isabella Venter
Institution Name: University of Western Cape
Active Member Count: 2
Allocation Start: 2021-01-04
Allocation End: 2021-07-14
Used Hours: 10250
Project Name: Analysis of irregular sequential data
Project Shortname: CSCI1383
Discipline Name: Computer Science
Research entitled : Analysis of irregular sequential data
Research field : Deep Learning of Artificial Intelligence
Short name : CSCI1383
Principal investigator : Professor Isabella Venter, Department of Computer Science, University of the Western Cape, South Africa (iventer@uwc.ac.za).
Researcher's details : Kudakwashe Dandajena, PhD Student in Computer Science at the University of the Western Cape, South Africa (3986658@myuwc.ac.za).
CHPC Contact : Kevin Colville (kcolville@csir.co.za)
Reporting Date : 5th of July 2021
This research has been utilising resources from the South African Centre for High Performance Computing (CHPC) since the 18th of December 2020 to date for computational processing and visualizations. The supercomputing environment provided us with graphical processing units (GPU), CUDA toolkit software environment for GPU deployment and Anaconda distribution software with Python and Jupyter Notebook, Keras, TensorFlow, Pandas and other libraries.
So far, we have utilised these resources to develop a systematic deep learning framework to resolve complex analysis challenges associated with discrete irregular sequential datasets. To design this framework, a process of combining context driven gated recurrent neural networks in the form of long short-term memory and gated recurrent units combined with bidirectional mechanisms and self-attention mechanisms was used. The framework was tested on sequential datasets such as the daily currency exchange rates which provide high levels of discrete irregular patterned environments. It improved regression performance accuracy by 38 % and stability performance by 24 % when compared to other baseline artefacts selected from several seminal research articles. The net effect of this study improved the understanding of the nature of deep learning artefacts. It also provided guiding principles for the development of a reliable and explainable deep learning toolbox composed of architectures, algorithms and models ranked using a multidimensional performance evaluation approach.
Principal Investigator: Prof Penny Govender
Institution Name: University of Johannesburg
Active Member Count: 7
Allocation Start: 2021-01-08
Allocation End: 2021-07-07
Used Hours: 439526
Project Name: Design of MOFs-based sorbent materials for capturing carbon dioxide
Project Shortname: CHEM0797
Discipline Name: Material Science
The computational chemistry research group at the university of Johannesburg is led by Prof Penny Govender. The research group currently focuses on material science, environmental sciences, reaction mechanisms and medicinal chemistry. Curently, Covid-19 related research has taken the centre stage for our medicinal chemistry and related research. The work is being performed both for predictive purposes (where experiments are modelled before being actually carried out in the lab) and for support purposes where computational results supports experimental results. All the simulations done by the group are carried out on the Lengau cluster using software such as Materials Studio, Gaussian, Schrodinger, and AMBER, all provided by the CHPC South Africa.
Principal Investigator: Prof Richard Betz
Institution Name: Nelson Mandela Metropolitan University
Active Member Count: 15
Allocation Start: 2021-01-08
Allocation End: 2021-07-07
Used Hours: 18499
Project Name: Artificial Photosynthesis
Project Shortname: CHEM0872
Discipline Name: Chemistry
The research group of Dr Betz at Nelson Mandela University does research on the field of artificial photosynthesis.
The latter process needs to be understood to enable "farming" and food production even in places where natural plant growth is not/no longer possible (e.g. because of climate change).
The goal shall be achieved by means of creating the smallest factory possible, i.e. "A molecule that can do this job".
Calculations and simulations help to assess certain pathways and steer the research.
Principal Investigator: Dr Collins Obuah
Institution Name: University of Ghana
Active Member Count: 3
Allocation Start: 2021-01-08
Allocation End: 2021-07-07
Used Hours: 48532
Project Name: Bioinorganic Chemistry
Project Shortname: CHEM1351
Discipline Name: Chemistry
This project is to look for a better catalyst for the oxidation of ethylene to vicinal diols. Currently the catalyst used for the process is OsO4. This is a good catalyst bad its expensive and very toxic. Therefore the need to search of an inexpensive and non-toxic catalyst for this process. Our computational investigation shows that catalysts of the form LReO3 have the potential to replace OsO4 as catalyst for the oxidation of ethylene.
The research is a computation studies therefore CHPC is the primary tool for the success of the project.
Principal Investigator: Prof Orde Munro
Institution Name: University of the Witwatersrand
Active Member Count: 6
Allocation Start: 2021-01-11
Allocation End: 2021-07-10
Used Hours: 135261
Project Name: Bioinorganic Chemistry
Project Shortname: CHEM1065
Discipline Name: Chemistry
BIOINORGANIC RESEARCH GROUP – WITS UNIVERSITY
The Bioinorganic Research Group at WITS University is led by Prof. Orde Munro, who is currently the SARChI Chair of Bioinorganic Chemistry. At present, the WITS group comprises three post-doctoral research fellows (Drs. Catherine Slabber, Zeynab Fakhar, and Angelique Blanckenberg), four PhD students, and two MSc students.
NATURE OF OUR WORK
The work mainly involves the design, synthesis, and testing of novel metallodrug candidates for chemotherapy. The compounds incorporate ligands that can facilitate and control the interaction of the metal ion with biomolecules such as DNA and/or proteins. Molecular simulations are used not only to design compounds with desirable biological behavior, but also to gain fundamental insights on their mechanisms of action and electronic structures. We are particularly interested in exploiting and controlling the reactivity of both endogenous and non-natural metal ions to kill drug-resistant pathogenic bacteria such as TB (Mycobacteria) and cancer cells.
WHY WE USE SIMULATIONS
Modern drug discovery uses in silico methods for a number of reasons, including screening large libraries of compounds for binding efficacy to enzyme targets. The enzyme targets are crucial proteins for the survival and replication of cancer cells, bacteria and viruses. Screening many thousands of compounds using computer programs is an efficient way to select a handful of compounds that can ostensibly bind to the enzyme target of interest because they have the correct shape and functional groups. The small library so generated may then be synthesized in the laboratory and tested against the pathogen of interest and the putative enzyme target. In this way, computational methods shorten the time required to find promising new compounds for further testing and development as the medicines of tomorrow. Finally, computer simulations offer important three dimensional details about how drug-like compounds inhibit their enzyme targets, thereby guiding optimization of the compound's structure for enhanced target affinity and, in principle, improved in vivo efficacy.
PROJECT PROGRESS
Over the last 2 years the project has produced 2 published papers in ISI-rated international journals with another 4 papers in the pipeline. Importantly, use of the CHPC has expanded to all students in the group from MSc to post-doctoral level. The PI has also commenced use of the CHPC resources after finding the time to learn how to use the system. Regarding the overall thrust of the work, the CHPC resources have supported efforts to discover novel soluble epoxide hydrolase inhibitors, delineate the mechanism of action of gold(III) quinoline-amide topoisomerase II inhibitors, and design a library of novel metallodrug candidates for SARS-CoV-2. The latter compounds are being synthesized for in vitro screening against the virus.
Principal Investigator: Prof Christopher_W Cleghorn
Institution Name: University of the Witwatersrand
Active Member Count: 3
Allocation Start: 2021-01-12
Allocation End: 2021-07-11
Used Hours: 45679
Project Name: Population Based Metaheuristics for Machine Learning
Project Shortname: CSCI1385
Discipline Name: Computer Science
The research undertaking in the research project at WITS, focused on the hybridization of machine learning and population based approaches. This area of research is fundamentally impactful as it can be applied in areas where gradient information is not readily available, which allows a broader degree of applicably than just classic machine learning. The CHPC has allowed for this important work to proceed without being unduly contained by computational resources, and hence has empowered the students to tackle computationally challenging tasks that are of Importance to the AI community as a whole.
Principal Investigator: Dr Moses Okpeku
Institution Name: University of KwaZulu-Natal
Active Member Count: 5
Allocation Start: 2021-01-12
Allocation End: 2021-07-11
Used Hours: 18472
Project Name: Genomics and Bioinformatics Group Westville Campus
Project Shortname: CBBI1185
Discipline Name: Bioinformatics
Using the CHPC resources has made research in my group cheaper, faster and led to much-desired success in vaccine design. We benefited from the CHPC beyond what our desktop computers could not do, we were able to use expensive software that we ordinarily could not access as a small group. continued support for keeping CHPC alive and running is highly.
Principal Investigator: Dr Leigh Johnson
Institution Name: University of Cape Town
Active Member Count: 3
Allocation Start: 2021-01-12
Allocation End: 2021-07-11
Used Hours: 199162
Project Name: MicroCOSM: Microsimulation for the Control of South African Morbidity and Mortality
Project Shortname: HEAL1049
Discipline Name: Health Sciences
The MicroCOSM project, based at the Centre for Infectious Disease Epidemiology and Research at the University of Cape Town, is a project to simulate the spread of infectious diseases, as well as the incidence of non-infectious diseases, in the South African population. By simulating the social and biological factors that contribute to disease transmission, we can better understand which sub-populations need to be targeted for special interventions. We can also evaluate the impact that various prevention and treatment programmes have had in South Africa to date, and evaluate the potential impact of new programmes. This simulation involves generating nationally representative samples of 20,000-40,000 South Africans and tracking them over their life course. Because this involves many individual-level calculations, the model requires substantial computing power, and the CHPC is therefore critical to the conduct of these simulations. We have recently used the model to assess strategies to reduce the future incidence of cervical cancer in South Africa, and to assess the impact of different HIV programmes on trends in HIV incidence.
Principal Investigator: Prof Andrei Kolesnikov
Institution Name: Tshwane University of Technology
Active Member Count: 2
Allocation Start: 2021-01-13
Allocation End: 2021-07-12
Used Hours: 89748
Project Name: Modeling and simulation of multiphase flows with application in processing industries
Project Shortname: MECH1386
Discipline Name: Chemical Engineering
M.Eng. Chemical Engineering students Alex SEYA and Justin MBWEBWE under supervision of Prof. Andrei KOLESNIKOV are involved in the modeling and simulation of plasma spheroidization and plasma spraying processes at Tshwane University of Technology. Plasma spheroidization process works by passing raw material—angular particles of metal or previously used powders—through high-energy plasma gas. The material melts and is reformed into spherical particles with few satellites (smaller particles that stick to larger ones); this results in a batch of powder that is more suitable for additive manufacturing. Plasma spraying is used to apply ceramic, cermet and metal coatings in several industry sectors. The powder is injected into plasma jet and accelerated and melted. The molten particles heat the substrate and form a coating. These coatings are designed to provide wear and corrosion protection, or to give functional properties such as thermal insulation or biocompatibility. To obtain quality products (spheroidized powders and coatings), insight into the links between operational parameters and product quality is required. CFD modeling and simulation helps to reveal such links and optimize the process. Multiphase modeling and simulation of unsteady turbulent plasma flows with high temperature and velocity gradients require fine computational mesh and small time steps, therefore, utilization of multi-core clusters with large memory is imperative in order to achieve results in reasonable time.
Principal Investigator: Prof Jacques Joubert
Institution Name: University of Western Cape
Active Member Count: 2
Allocation Start: 2021-01-13
Allocation End: 2021-07-12
Used Hours: 145508
Project Name: Polycyclic cage compounds as multifunctional neuroprotective agents
Project Shortname: CHEM1173
Discipline Name: Chemistry
The UWC drug design and discovery group was able to use CHPC tools to generate additional infomation on polycyclic cage-like compounds. We are exploring these molecules as possible scaffolds that can be used to develop more effective neuroprotective agents. We are now taking this project forward and we will now study the ability of these molecules to bind to certain targets using molecular modelling and quantum chemical analysis.
Principal Investigator: Prof Ken Craig
Institution Name: University of Pretoria
Active Member Count: 3
Allocation Start: 2021-01-12
Allocation End: 2021-07-11
Used Hours: 241324
Project Name: CFD modelling of jet impingement with phase change
Project Shortname: MECH1296
Discipline Name: Computational Mechanics
Jet impingement heat transfer with boiling is a promising new heat transfer method for use in power electronics cooling. Most of the current research on this topic in literature is experimental, leaving a gap for computational studies to further investigate the relevant influencing parameters. Jet impingement flow is difficult to simulate because of its transient nature and multiple flow regimes involved (from laminar to turbulent). Previous research by Prof Ken Craig's group has developed techniques based on Large Eddy Simulation to evaluate the effect of round jets with and without swirl on solar receiver heat transfer surfaces. This program extends that work to having phase change as well in the form of boiling. The latent heat absorbed by boiling works in addition to the enhance convection provided by the jet. Various case studies are being developed for application in the electronics cooling industry and the approach has been validated using test cases from literature.
Principal Investigator: Prof Ken Craig
Institution Name: University of Pretoria
Active Member Count: 4
Allocation Start: 2021-01-12
Allocation End: 2021-07-11
Used Hours: 764434
Project Name: CFD Modelling of Concentrated Solar Power Applications
Project Shortname: MECH0873
Discipline Name: Computational Mechanics
Central tower solar power plants are the most promising concentrated solar with research into new receivers and heat transfer fluids and collectors (heliostats) being ongoing world-wide. Researchers at the University of Pretoria's Clean Energy Research group under the leadership of Prof Ken Craig are using the resources at the CHPC to speed up calculations of both heliostat analysis (involving wind loads and fluid-structure interaction) as well as those required for a new point-focus central receiver design being developed. For heliostats, knowledge has been gained in predicting the peak loads that these reflectors will experience during all orientations and wind conditions. Ongoing work is focused on further reducing the cost of these Computational Fluid Dynamics (CFD) hybrid Large Eddy Simulation (LES):Reynolds-Averaged Navier-Stokes models to make them feasible on smaller-scale computers. This project has been ongoing since 2013 with the involvement of CHPC resources since 2016. The project has already delivered three masters students (Dawie Marais, Joshua Wolmarans and Derwalt Erasmus) with Jesse Quick nearing completion. PhD student Pierre Poulain is also nearing the end of this investigation into LES modelling of the atmospheric boundary layer and determining peak loads on solar collectors. For central receivers, work is ongoing to enhance heat transfer. Of specific interest here is the use of jet impingement, especially in the transient mode through either passive or active excitation. This phenomenon of heat transfer enhancement, although conceptually simple, it very hard to model accurately, and here the resources of the CHPC are invaluable, especially as LES is required to model the transient and 3D nature of swirling jet, the focus of the work of Jesse Quick. CHPC resources were also used to investigate the thermal performance solar dish receivers by accurate simulation of heat losses due to dish orientation and wind speed.
Principal Investigator: Dr Samuel Egieyeh
Institution Name: University of Western Cape
Active Member Count: 14
Allocation Start: 2021-01-14
Allocation End: 2021-07-13
Used Hours: 4838
Project Name: Computational (Cheminformatic and Bioinformatic) Drug Discovery, Design and Development for Infectious Diseases
Project Shortname: CBBI1212
Discipline Name: Health Sciences
The Computational Pharmacology
and Cheminformatics Research Group operates from the School of Pharmacy, University of Western Cape. Our research focus on computational drug discovery and
design and data science for predictive drug development and precision medicine for infectious diseases.
In drug discovery, the journey from "hits" to "drug candidates"is tedious, long and expensive. A high quality drug candidate must exhibit a balance of many properties, including potency and safety/toxicity. Hence
a multi-parameter optimization strategy is required. We use the CHPC platform for rigorous, high-throughput data analytics, data mining and computational modelling
to achieve our aim of developing potential "likely to succeed" drug candidates for infectious diseases.
Principal Investigator: Prof Richard Tia
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 31
Allocation Start: 2021-01-14
Allocation End: 2021-07-13
Used Hours: 3691909
Project Name: Computational Chemistry and Materials Science
Project Shortname: CHEM1044
Discipline Name: Chemistry
The work we do here with CHPC resources address critical aspects of the chemical industry, renewable energy and mitigation of climate change.
The work we do fall under the following areas:
1. Computational studies on the synthetic pathways relevant to the pharmaceutical industry. We study the mechanisms of chemical reactions underlying the synthesis of pharmaceutically-relevant molecules. The information from these studies are useful in guiding medicinal chemists in the design of targeted molecules for pharmaceutical applications.
2. Computational studies on homogeneous catalysis used in the chemical industry. These studies afford important information on the rationale design of chemicals for fuels, plastics etc
3. Computer-aided design of materials for applications in solar energy generation, green-gas mitigation etc. The atomic- and molecular-level mechanistic insight gained from these studies are useful for the design of novel catalysts for the conversion of CO2 and other green-house gases as well as stranded gases from the petroleum refining into useful fuels. This addresses the problem of global warming as well as providing an environmentally friendly alternative sources of fuel.
The research objectives described here are very computationally-demanding and using the CHPC resources is highly beneficial. Without that, we would not be able to carry out some of these calculations.
Principal Investigator: Mr Reshendren Naidoo
Institution Name: University of the Witwatersrand
Active Member Count: 2
Allocation Start: 2021-01-12
Allocation End: 2021-07-11
Used Hours: 159906
Project Name: Influence of Gas Phase Kinetics on a Pulverised Coal Flame
Project Shortname: MECH1298
Discipline Name: Computational Mechanics
The combustion of coal particles is a complex process and there are several physical processes at play. These processes are in turn highly dependent on the quality of coal being combusted and hence there must be direct link between the numerical methods employed and the true heterogeneous nature of coal. One of these facet is the way combustible material is released from coal particles. The chemical mechanisms that are used to predict the flame characteristics are important to understand the ignition/stability and emissions profile of a given combustion regime. Hence it is important to study these chemical mechanisms.
Principal Investigator: Mr Ernest Opoku
Institution Name: Nesvard Institute of Molecular Sciences, Ghana
Active Member Count: 7
Allocation Start: 2021-01-18
Allocation End: 2021-07-17
Used Hours: 116347
Project Name: Molecular Quantum Chemistry
Project Shortname: CHEM1352
Discipline Name: Chemistry
The Molecular Quantum Chemistry Research Group is affiliated with the Department of Basic and Computational Sciences, Koachie Health Systems, Accra, Ghana.
We are a team of young and gifted conscious Africans, who are voluntarily using our skills and expertise to promote Kaochie Health Systems (KHS) through research to help elevate the reputation of Ghana and Africa on the map of scientific research. Our motivation for what we do is that "we want to be part of the patriotic Africans to build Africa for Africans".
Modern chemical and pharmaceutical industries require reactions that are energy efficient and selective towards formation of various relevant compounds. Thus, synthetic designs require the ability to carry out multiple chemical transformations in fewer steps to achieve higher synthetic efficiency. To achieve this, a thorough understanding of the intrinsic steps involved in such chemical transformations and the role of catalysts are crucial. Such information is efficiently obtained using high-performance computers to do extremely demanding calculations. Africa needs scientists with high performance computing expertise coupled with multidisciplinary scientific background to provide a complement to present and future African pharmaceutical and chemical industries. We hope that our objectives will be interesting to young African scientists to voluntarily come on board with their expertise for Africa's overall progress. Multiple projects are currenting ongoing using the CHPC computational resources. Manuscripts are under review and many more are in preparation.
Principal Investigator: Prof Mwadham Kabanda
Institution Name: University of Venda
Active Member Count: 2
Allocation Start: 2021-01-18
Allocation End: 2021-07-17
Used Hours: 368658
Project Name: Reaction mechanism for atmospheric relevant molecules
Project Shortname: CHEM1161
Discipline Name: Chemistry
The group is currently composed of the group leader Prof MM Kabanda. The research aim of the group is the investigation of the reaction mechanisms and properties for selected molecules that possess usefulness both in the atmosphere and within biological systems.
The research is being done because of the challenges posed by radical species both in the atmopshere and within the biological systems. Understanding their reaction mechanism and reaction rate patterns would help a long way to understand the patterns of other reactions that may not be easily modelled.
The investigation is purely theoretical and depends on computational resources at CHPC. We are grateful for the computational resources and softwires that are currently being provided for public use.
The project is proceeding well.
Principal Investigator: Prof Marelie Davel
Institution Name: North-West University
Active Member Count: 11
Allocation Start: 2021-01-18
Allocation End: 2021-07-29
Used Hours: 273930
Project Name: DNN Analysis
Project Shortname: CSCI1191
Discipline Name: Computer Science
MuST is a research group at the Faculty of Engineering at NWU whose work is focused on the study of machine learning, with a strong track record of applying these in speech processing applications. We have a specific interest in the theory and application of deep learning techniques. Our application domains are varied, ranging from speech and language processing, to space weather prediction, to industrial applications of deep learning.
In the past decade, the field of Deep Neural Networks (DNNs) has brought renewed energy and focus to AI, through a series of remarkable breakthroughs in fields as diverse as speech recognition, board games and self-driving cars. In these and other applications, DNN systems have reached previously unknown levels of accuracy, making human-level performance a distinct possibility and thus suggesting novel insights on the mind-matter problem.
The successes of DNN systems have inspired much research into better algorithms, novel applications and a better understanding of DNNs. The MuST group is involved in all these aspects of DNN research. For example, we are using DNNs and word embeddings to develop better language models for under-resourced languages; these models can be used in tasks such as speech recognition and machine translation. We are also using DNNs to handle poor quality audio in speech and speaker recognition systems better, probe the processes at play during solar flare eruptions, and even optimising the design process of airfoil shape with some of our industry partners. We balance these applications with theoretical work focussed on understanding and characterising generalisation in the context of deep learning.
MuST hosts the CAIR Deep Learning group, a node of the South African National Centre for Artificial Intelligence Research (CAIR), an initiative aimed at 1) developing world-class AI research leadership and teaching capability in South Africa; 2) supporting sustainable and effective socio-economic development through consolidated, applied AI research initiatives; and 3) building an accredited national and international AI research network that promotes AI research and technology in South Africa.
MuST has a long track record with regards to R&D project delivery for international NGOs, international commercial clients, international government agencies, and local government. The group consists of senior researchers and postgraduate students who work side by side to investigate some of the interesting, elusive topics in AI.
MuST seeks collaboration opportunities with partners that supplement our skillset. Notable current collaborators include Saigen, a South African start-up company that that develops and deploys deep neural network based speech-to-text solutions for localised commercial applications, and the South African National Space Agency (SANSA), with whom we explore deep learning in geospace applications.
Principal Investigator: Prof Jeanet Conradie
Institution Name: University of the Free State
Active Member Count: 5
Allocation Start: 2021-01-19
Allocation End: 2021-07-18
Used Hours: 1266361
Project Name: Computational chemistry of transition metal complexes
Project Shortname: CHEM0947
Discipline Name: Chemistry
This project is concerned with the computational chemistry of transition metal complexes, with special focus on the application of computational chemistry in determining the structure and energy of transition metal complexes, transition states and reaction-intermediates. This is reflected in the outputs of my research group and collaborators during the period of report. The outputs gave a better understanding of experimental observation and of the factors influencing the reactivity of metal complexes, in order to streamline them for use in different applications, for example in catalysis or energy sources (batteries and solar cells), nonlinear optics, medical applications (eg HIV and anti-leukemia) and antimicrobial activity (anti- bacterial and anti-fungal).
Principal Investigator: Mr Muaaz Bhamjee
Institution Name: University of Johannesburg
Active Member Count: 6
Allocation Start: 2021-01-19
Allocation End: 2021-07-18
Used Hours: 30979
Project Name: Modelling of Multiphase Flow in Process Equipment
Project Shortname: MECH0837
Discipline Name: Computational Mechanics
he research group is from the Department of Mechanical Engineering Science at the University of Johannesburg. The primary area of interest in the research is the use of coupled computational fluid dynamics and computational granular dynamics models to investigate the complex multiphase interaction in process equipment.
The models are based on multiple continuum and meso-scopic based approaches which are couple with models such as the Discrete Element Method, Immersed Boundary Method and Eulerian Dense Discrete Phase Models. Both commercial and open-source software has been used in the research. Primarily the research has been focused on
hydrocyclones. However, the research is moving onto fluidised beds, spiral concentrators, highly concentrated slurry flows and gas cyclones. The models are benchmarked accuracy) against experimental measurements as well as for computational efficiency. Accurate and computationally efficient modelling of such phenomena is vital to the design of hydrocyclones which is crucial to the mining industry. The research group has also begun to use this platform to model fluid dynamics in biomedical applications. The research group has produced numerous publications and has supported the research of four post-graduate students. Research into the computational efficiency of the models is continuing with the aid of the Centre for High Performance Computing (CHPC) facilities. Due to the complex physics and fidelity required in the problems tackled by the research group large computational resources are required. The CHPC provides a computational platform that is vital to the ongoing work and success of the research group.
Principal Investigator: Prof Thomas Niesler
Institution Name: Stellenbosch University
Active Member Count: 10
Allocation Start: 2021-01-19
Allocation End: 2021-07-18
Used Hours: 362608
Project Name: Automatic Speech Recognition for Under-resourced Languages
Project Shortname: CSCI1059
Discipline Name: Electrical Engineering
Most languages spoken in Sub-Saharan Africa are not technologically developed. They do not have, for example, the carefully prepared collections of speech and text that are required for the development of automatic speech recognition and associated technologies. The Digital Signal Processing Laboratory, which is part of the Department of Electrical and Electronic Engineering at Stellenbosch University, is focussing its research efforts on developing methods that allow the development of automatic speech recognisers from much smaller datasets - small enough to make their compilation feasible for the languages concerned. The developed techniques are equally applicable to South African languages, which also remain largely under resourced. For example, the group has develop the world's first pentalingual speech recognition system that is able to switch back and fourth between five South African languages (isiXhosa, isiZulu, Sesotho, Setswana and English). Such code-switching, where a speaker alternates between languages in the same utterance, is common in spontaneous South African Speech and in other multilingual situations. The development if these speech recognition systems requires the extensive training an optimisation of existing and new neural network architectures. A resource like the CHPC is essential for such computationally demanding work. The programme is progressing steadily, and has gained international attention.
Principal Investigator: Prof Eric van Steen
Institution Name: University of Cape Town
Active Member Count: 7
Allocation Start: 2021-01-20
Allocation End: 2021-07-19
Used Hours: 871138
Project Name: Lateral interactions on surfaces
Project Shortname: CHEM0780
Discipline Name: Chemistry
Researchers at the Catalysis Institute in the Department of Chemical Engineering at the University of Cape Town have developed a novel method for the selective oxidation of methane over a heterogeneous catalyst. This will open a new, environmentally more benign route for the conversion of natural gas into chemicals. Water, and in particular the interaction of water with species adsorbed on the surface play a key role in this process. Detailed molecular modelling has provided the initial idea for this process, which was further backed up by large scale exploration on the interaction of water with platinum-based surfaces.
Principal Investigator: Prof Linky Makgahlela
Institution Name: Agricultural Research Council
Active Member Count: 6
Allocation Start: 2021-01-20
Allocation End: 2021-07-29
Used Hours: 66798
Project Name: Beef and Dairy Genomics
Project Shortname: CBBI1138
Discipline Name: Bioinformatics
The ARC-Animal Production, Animal Breeding and Genetics unit is based in Irene, Pretoria. Our research activities aim to develop knowledge on the animal genetic resources of South Africa, and apply traditional and genomics technologies for improved climate-smart production in the livestock industry while preserving animal biodiversity. The growing global population threatens food security, and half of this growth is expected in Africa. Climate change will reduce available natural resources needed for agriculture. We need to produce 100% more food, than today, in the next 30 years, and with limited natural resources. Genomic technological innovations promises to deliver genetic diversity, efficiency, product quality, food safety and public health, animal health and welfare and environment, all interlinked. Current research develops strategies for characterization, conservation and utilization of indigenous animal (cattle, sheep & goats) genetic resources. These are adapted genotypes that will survive global warming. Combine advance genomic technology data (DNA markers and whole genome sequence) and traditional animal pedigrees and longitudinal performance data, and develop statistical models for national genetic evaluations estimated breeding values. The use of genomically enhanced breeding values will speed up the rate of identifying the breeding stock and food production in cattle, sheep and goat sectors.
All research activities are carried out in collaboration with partners e.g. SA Universities, Commodity groups and Government departments. Post-graduate students are linked to the research as the bloodline for capacity building. Beef genomic research results in inbreeding and imputation (MSc. and PhD degrees awarded) have been disseminated to the cattle breeders association to facilitate uptake. Dairy genomic research is ongoing, and all projects utilizing whole genome sequences are ongoing.
Principal Investigator: Prof Fernando Albericio
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-01-20
Allocation End: 2021-07-29
Used Hours: 26659
Project Name: Peptide chemistry and Organic chemistry
Project Shortname: CHEM1102
Discipline Name: Chemistry
We are Peptide Science Group at UKZN Westville campus. The project was brought up with an intend to learn computational chemistry. The project involves synthesis of several organic molecules towards application in oligonucleotide chemistry. With the help of computational tools we aim to predict the reaction mechanism. Since CHPC provides the server which speeds up the calculations and helps in achieving the preliminary results. The project so far had been very promising in its preliminary stage. We expect to be going ahead with these results for further analysis towards prediction of reaction mechanism.
Principal Investigator: Prof Thomas Scriba
Institution Name: University of Cape Town
Active Member Count: 3
Allocation Start: 2021-01-22
Allocation End: 2021-07-21
Used Hours: 29929
Project Name: Tuberculosis vaccine and biomarker development
Project Shortname: HEAL1390
Discipline Name: Health Sciences
The South African Tuberculosis Vaccine Initiative (SATVI) is the largest organization dedicated to clinical TB vaccine development. Comprised of a clinical immunology team of scientists, post-doctoral fellows and postgraduate students, as a field site team based in Worcester, SATVI has completed more than 30 phase 1/2/2b clinical trials of novel TB vaccines. A major additional focus aims to identify and validate human immune correlates of risk of TB, and of protection against of TB. For example, we are working on several projects that seek to identify correlates of protection using samples collected in two major clinical trials with efficacy against TB (one of BCG revaccination and the other of the M72:ASO1E vaccine), which we at SATVI contributed to substantially. The SATVI clinical immunology team has gained valuable experience in clinical study design, systems immunology, and computational approaches to generate and parse high-dimensional data such as transcriptomics and proteomics, TCR sequencing, multiparameter flow cytometry and CyTOF, as well as cellular function. Our latest work, which is currently under, seeks to analyse whole blood transcriptome datasets to understand asymptomatic TB, a very significant issue that is very poorly understood.
Principal Investigator: Dr Kevin Lobb
Institution Name: Rhodes University
Active Member Count: 13
Allocation Start: 2021-01-25
Allocation End: 2021-07-24
Used Hours: 1202983
Project Name: Computational Mechanistic Chemistry and Drug Discovery
Project Shortname: CHEM0802
Discipline Name: Chemistry
This group has two main thrusts. One is understanding of mechanism and of material properties and two students are working on this. B. Isamura is looking in depth at a single reaction type and how careful exploration of forces during reaction may be used to classify this reaction type into subtypes. F. Koohyar is looking at solvation of materials and the properties associated.
On the other hand we explore drug discovery and biological mechanism. So M Mbunge, S Zabo and Z Sanusi are exploring Covid-19, the mechanism of action of the protease and what can be done to inhibit this protease using covalent docking.
Principal Investigator: Mr Randall Paton
Institution Name: University of the Witwatersrand
Active Member Count: 7
Allocation Start: 2021-01-25
Allocation End: 2021-07-24
Used Hours: 758105
Project Name: Compressible Gas Dynamics
Project Shortname: MECH0847
Discipline Name: Other
The work group led by Dr Randall Paton is looking at the effects of rapid, arbitrary acceleration, and specialised geometry, on compressible aerodynamics continues. This work is also being extended to examine some effects in space propulsion, and isentropic interactions of compressible flows. There are also plans to implement machine learning techniques to expand computational modelling capabilities.
The CHPC is an integral part of this work. It allows for the parametric simulation of these flow fields, which are often impossible to produce experimentally. The current cohort of several Masters and two PhD students all make use of this incredible facility (expertly managed by its tirelessly-dedicated staff) to expand the boundaries of knowledge, leading graduates from the group of all ethnic backrounds and sexes to careers both locally and abroad.
Principal Investigator: Prof Martin van Eldik
Institution Name: North-West University
Active Member Count: 3
Allocation Start: 2021-01-26
Allocation End: 2021-09-30
Used Hours: 271877
Project Name: Thermal-Fluid modelling
Project Shortname: MECH1247
Discipline Name: Computational Mechanics
The Thermal-Fluid Modelling research group, is situated in the Faculty of Engineering at the North-West University. Students using the CHPC currently consists of 1 Masters student and 1 PhD. All students are using Star-CCM+ for their studies. The research group produced one masters student in 2020 that investigated the thermal fluid modelling of a titanium helically coiled fluted tube. The current Masters student is conducting an aeroacoustic analysis of a glider plane component and how to identify the source and replicate it. One of the PhD students is investigating the CFD (computational fluid dynamics) modelling of a sinter mixing drum to improve the production throughput. The CHPC makes it possible for the students to be able to complete their work on time, which otherwise would not have been able to.
Principal Investigator: Dr Jo-Anne de la Mare
Institution Name: Rhodes University
Active Member Count: 1
Allocation Start: 2021-01-26
Allocation End: 2021-07-25
Used Hours: 10821
Project Name: Characterization of novel inhibitors for triple negative breast cancer
Project Shortname: HEAL1391
Discipline Name: Other
Despite the disruptions due to load shedding, the service has been very good. The CHPC resources is highly beneficial to my research, allowing us to perform comprehensive in silico studies in terms of modelling interactions between hit compounds and their proposed target that strengthen our hypotheses that the novel compounds indeed target specific oncogenic proteins in cancer cells. These kinds of analyses would not be possible using available resources which allow for only static docking of compounds to the proposed targets. On the other hand, the molecular dynamics analysis made possible by the CHPC resources, provide far greater insights into these putative interactions.
Principal Investigator: Dr Abdulrafiu Raji
Institution Name: University of South Africa
Active Member Count: 12
Allocation Start: 2021-01-26
Allocation End: 2021-08-12
Used Hours: 4135421
Project Name: Structural, electronic, magnetic properties and anisotropy energy in some metallic and non-metallic heterostructures
Project Shortname: MATS0988
Discipline Name: Material Science
The research work is a collaboration between Dr. Raji (Physics Department, UNISA, South Africa) and the group of Dr. Brice Rodrigue Malonda (Marien Ngouabi University, Republic of Congo). The focus of the research is numerical studies of electronic, optical, transport and magnetic properties of selected two-dimensional (2D) and three-dimensional (3D) solid materials for potential applications in high-capacity data storage, catalysis and renewable energy. The study employs density-functional theory (DFT) investigate the atomic-level properties of these materials. We aim to modify the pristine properties of selected solids through simple defect engineering of their structure, alloying and interface engineering.
The last two decades have witnessed efforts at discovering materials with novel properties for specific applications such as in electronic and magnetic devices. The discovery of new materials are sometimes accompanied by previously unknown physical, chemical or electronic processes which necessitate the use of advanced theoretical approaches to gain deeper understanding of these processes. One of the aims of our research therefore, is to discover novel materials via computational methods and to exploit the unique properties of these materials for potential technological applications. Ultimately, our research is theory-led discovery of novel materials.
The numerical implementation of DFT is computational intensive, requiring several computational hours, large data storage and memory requirements, beyond the capacity of ordinary table-top computers. Therefore, high-capacity computer clusters such as the one provided by the Center for High Performance Computing (CHPC) is sine-qua-non for the research. There are about 10 postgraduate students of various nationalities working in various aspects of the project. This research, as well as the CHPC, has enabled collaboration between South Africa based academic researcher and colleagues in Congo, Mexico and Italy. There have been research outputs in form of publications, and the scope of the project will guarantee regular postgraduate students training who are able to undertake cutting-edge research and contribute to scientific development of South Africa, DRC and the Africa continent.
Principal Investigator: Prof Charalampos (Haris) Skokos
Institution Name: University of Cape Town
Active Member Count: 5
Allocation Start: 2021-01-27
Allocation End: 2021-07-26
Used Hours: 1224889
Project Name: Chaotic behavior of Hamiltonian systems
Project Shortname: CSCI1007
Discipline Name: Applied and Computational Mathematics
Using modern numerical techniques of Nonlinear Dynamics and Chaos Theory we investigate in a unified mathematical way the behavior of models describing the energy transport in disordered and granular media, as well as the properties of new elastic materials like graphene and molecules like the DNA. This research is performed by members of the 'Nonlinear Dynamics and Chaos Group' in the Department of Mathematics and Applied Mathematics at the University of Cape Town. The performed investigations constitute an innovative combination of Applied Mathematics, Chaos Theory, Hamiltonian Dynamics and Nonlinear Lattice Dynamics. Their outcomes will provide answers to some fundamental questions about the effect of chaos on the energy transport in disordered and granular media, and on the behavior of graphene and DNA. Some of the questions we try to address are: Does the presence of impurities and nonlinearities enhance or suppress the propagation of energy (e.g. heat) in solids, of light in crystals, and of vibrations in granular material? How does the ratio of the different base pairs in DNA chains affect their structural stability and the temperature at which the double-stranded DNA breaks to single-stranded DNA?
Principal Investigator: Prof Juliet Pulliam
Institution Name: Stellenbosch University
Active Member Count: 18
Allocation Start: 2021-01-27
Allocation End: 2021-07-26
Used Hours: 18770
Project Name: South African Centre for Epidemiological Modelling and Analysis
Project Shortname: CBBI1106
Discipline Name: Other
SACEMA is a DSI-NRF Centre of Excellence focused on Epidemiological Modelling and Analysis, hosted at Stellenbosch University. We use computational approaches to address questions of public health relevance to South Africa and the African continent. We use the CHPC primarily to run stochastic simulations and for simulation-based validation of novel inference methods.
Principal Investigator: Dr Emmanuel Sakala
Institution Name: Council for Geoscience
Active Member Count: 17
Allocation Start: 2021-01-25
Allocation End: 2021-09-02
Used Hours: 4993
Project Name: Application of high performance computing at the Council for Geoscience
Project Shortname: ERTH1227
Discipline Name: Earth Sciences
The Council for Geoscience programme involves running high computing packages for geophysical modelling, seismological data processing, geological modelling and any other packages which currently are not optimum for a desktop computing. Research and development of the various geoscience fields (geophysics, geology, geochemistry, hydrogeology) are covered in this programme.
The prgramme is progressing very well and we are in the third year of its existence.
Principal Investigator: Prof Robin Emsley
Institution Name: Stellenbosch University
Active Member Count: 7
Allocation Start: 2021-01-27
Allocation End: 2021-07-26
Used Hours: 266548
Project Name: EONCKS
Project Shortname: CBBI1064
Discipline Name: Health Sciences
The Psychosis Research Team of the Department of Psychiatry at Stellenbosch University continues to publish research aimed at improving the service delivery and understanding of psychosis-related disorders with a primary focus on Schizophrenia. Our work has addressed some important questions regarding which factors impact treatment outcomes which speak to the overall question of how we can improve the treatment of schizophrenia. Some of our most exciting work has been in the field of neuroimaging, utilising MRI scans to investigate potential neurobiological differences underpinning the disorder. These findings form part of the building blocks toward understanding the origins of psychosis and help the scientific community identify potential targets for treatment. Not only does that put us at the cutting edge of psychosis research, it also highlights the heights that can be achieved in what is generally considered a resource-constrained environment. Neuroimaging research is an expensive and time-intensive undertaking and it is imperative that the data be utilized to the utmost extent. Use of the CHPC Lengau cluster has been instrumental in this regard. We collected our MRI data between 2007 and 2018 and have since processed these scans with seven distinct pipelines, each yielding important new information that has been published in several high impact scientific journals. Manual processing of our data would have taken months per pipeline, whereas parallel processing on the cluster has reduced that time to weeks. In addition, we are able to run multiple processing pipelines concurrently. Use of the Lengau cluster is effectively the secret to our Research Team's ongoing ability to participate in a meaningful way in the scientific dialogue along with our collaborators in developed countries with greater access to resources. We remain committed to providing the best quality publications and to contribute to the international dialogue on the origin and treatment of psychosis.
Principal Investigator: Prof Tony Ford
Institution Name: University of KwaZulu-Natal
Active Member Count: 3
Allocation Start: 2021-01-27
Allocation End: 2021-07-26
Used Hours: 155886
Project Name: Computational studies of the properties of molecular complexes
Project Shortname: CHEM1134
Discipline Name: Chemistry
This research group is led by Emeritus Professor Tony Ford of the University of KwaZulu-Natal. It consists of three local colleagues, Professor Catharine Esterhuysen from the University of Stellenbosch, Dr Emmanuel Damoyi of Mangosutho University of Technology, and Professor Gert Kruger (UKZN), and an international collaborator, Professor Ponnadurai Ramasami of the University of Mauritius. Studies of the properties of some novel molecular complexes continue. Most recently the types of complexes which have attracted attention have been the halogen-bonded, chalcogen-bonded and triel-bonded families, which are all characterized by the presence of non-covalent bonds. This type of bonding is a recent addition to the range of interactions which are known to be responsible for the stabilities of such weak complexes, and information on such weak interactions can be very useful in understanding the more common types of bonding found in more familiar molecules. The work involves the use of a well-known standard computer program, Gaussian-16, which requires large allocations of computer time and memory. This requirement can only be met by an organization such as the CHPC which has the necessary computer infrastructure to make this type of work feasible.
Principal Investigator: Dr Ruben Cloete
Institution Name: University of Western Cape
Active Member Count: 8
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 712177
Project Name: HIV-1C integrase drug resistance
Project Shortname: CBBI1154
Discipline Name: Bioinformatics
The research group of Dr Ruben Cloete is based at the South African National Bioinformatics Institute, University of the Western Cape. The work in my group is primarily focussed on molecular modelling and drug design. Here we focus on protein structure prediction, molecular docking and simulation studies of protein-drug, protein-protein systems. Our research efforts is in understanding HIV-1 drug resistance, identifying novel drugs to treat drug resistant Tuberculosis and the prioritization of novel genes associated with Parkinson's disease in South African families. This work has led to the identification of new drugs to treat Tuberculosis. Furthermore, ongoing work might also contribute to the understanding of the development of Parkinson's disease and the better treatment of HIV-1 infected individuals within South Africa. For this to become a reality requires the use of structural computational methods to understand the binding of drugs to the molecular machines called proteins. Therefore, large scale computing resources are required to run large protein systems. Currently we are running simulations for several protein-DNA-drug complexes on the CHPC cluster We may even require more resources.
Principal Investigator: Dr Steven Hussey
Institution Name: University of Pretoria
Active Member Count: 2
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 242444
Project Name: Forestry Molecular Genetics - transcriptional and epigenetic regulation of wood formation
Project Shortname: CBBI1010
Discipline Name: Bioinformatics
The Forest Molecular Genetics Programme is a joint venture of the University of Pretoria and South African forestry industry partners aimed at developing biotechnology applications for tree improvement. It focuses on the genetic basis of tree growth, wood formation and defence against pests and pathogens. The SA forestry industry produces over 20 million tons of wood per year for a wide array of renewable products such as timber, pulp, paper, packaging, cellulose, textiles, pharmaceuticals and food additives. Increasingly, there is interest in using trees as biorefineries, i.e. energy-efficient production platforms for high-value biopolymers and biochemicals. Timber from genetically improved plantations can also be processed into advanced building materials for sustainable housing and construction. In part, the Programme aims to understand the biology of wood formation and how it is regulated at various levels. For example, the metabolism and incorporation of carbon-rich compounds formed from photosynthesis into cellular structures such as wood, and the role of networks of genes coordinating the activation of genes at the correct stages of growth and wood formation. An important part in understanding these process is the use of DNA sequencing technologies that produce data requiring high-performance computation to analyze it. These tools were central to decoding the genome or DNA sequence of Eucalyptus grandis, an important fast-growing tree. Our Programme has most recently relied on the Centre for High Performance Computing to understand how cellular organelles participate in wood formation by analysing gene expression patterns associated with the poorly understood plastid organelles found in wood, and implementing novel transfer machine learning approaches to infer gene regulatory networks underlying wood formation in Eucalyptus based on machine learning models built on thousands of gene expression and gene-gene interaction datasets from the model plant Arabidopsis.
Principal Investigator: Dr Holliness Nose
Institution Name: Technical University of Kenya, Nairobi, Kenya
Active Member Count: 1
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 149298
Project Name: Inorganic Computational Chemistry
Project Shortname: CHEM1003
Discipline Name: Chemistry
Our current project involves preparation of photosensitizers for water purification. Water treatment strategies commonly adopted include chlorination, distillation, UV-radiation, boiling and reverse osmosis. While these methods have proved to be effective in water decontamination, they have not been completely efficient in treatment of microbe polluted water. This is attributed to emergence of anti-microbial resistant (AMR) microorganisms such as Methicillin-resistant Staphylococcus aureus and Amphotericin-resistant Candida albican. AMR pathogens find their way into drinking water, causing waterborne diseases. Photodynamic antimicrobial therapy (PACT) provides an alternative and viable method of water disinfections, especially in addressing the anti-microbial resistant bacteria. PACT makes use of a photosensitizer, which when localized in microorganisms can be activated by irradiating it with light of appropriate wavelength to generate a reactive oxygen species (ROS). This ROS is able to destroy or affect biological structures such as cell wall material, nucleic acids, peptides and lipids therefore leading to cytotoxicity. Therefore, this method allows the photosensitizer to bind to the bacterial or fungal cell wall, inhibiting cell growth, with no possibility of recovery or development of resistance against it. My group intends to design and develop the photosensitizers computationally, followed by synthesis of the metal complexes and finally carry-out biological assays to examine their cytotoxicity. This project therefore relies on CHPC heavily as we need to design the photosensitizers theoretically before we go to the laboratory to synthesize them and test their biological activities. This project began on June 1st 2019 and and is currently ongoing. But without the CHPC facilities, we are unable to do this type of research. The project is proceeding on well.
Principal Investigator: Dr Adeola Joseph Oyenubi
Institution Name: University of the Witwatersrand
Active Member Count: 1
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 221064
Project Name: Optimizing balance in for Impact Evaluation
Project Shortname: ECON1213
Discipline Name: Economics and Finance
This research is conducted by Dr Oyenubi a Senior Lecturer at the University of the Witwatersrand. The research paper for this cycle looks at the use of multivariate balance in estimating causal impact under matching methods. This allows for the comparison of like-with-like in the multivariate sense. Earlier paper under this programme has shown that this is important for bias and efficiency of matching estimators. The analysis involve loops of up to 1000 and the CHPC cluster reduces the computation time considerably.
Principal Investigator: Dr Krishna Govender
Institution Name: University of Johannesburg
Active Member Count: 8
Allocation Start: 2021-01-27
Allocation End: 2021-07-26
Used Hours: 251469
Project Name: Computational approaches to design novel anticancer agents
Project Shortname: CHEM0792
Discipline Name: Chemistry
The access to CHPC allows me to provide much needed resources to my international collaborators that don't have access to such facilities. This in turn allows them to complete projects within reasonable times and permits students to complete their postgraduate studies within the University allocated time frames.
The CHPC resources also allow current projects that I am doing with local South African Universities to be completed timelessly.
Without the CHPC none of the work that I have done up to this point would be possible. Access to the resources has resulted in the publication of 30 high impact papers over the last 5 years.
Principal Investigator: Dr David Pearton
Institution Name: Oceanographic Research Institute (ORI)
Active Member Count: 4
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 10052
Project Name: Responses of South African Coral Reefs to Climate Change
Project Shortname: CBBI1072
Discipline Name: Environmental Sciences
The Oceanographic Research Institute (ORI), part of the South African Association for Marine Biological Research (SAAMBR) has been involved in researching and protecting South Africa's marine biodiversity since 1959. One of ORI's flagship programmes is the Coral Reef programme which has conducted pioneering research on the coral reefs of South Africa and the Western Indian Ocean.
Coral reefs all over the world are under significant pressure from a wide variety of sources including global changes such as warming oceans and ocean acidification as well as local threats such as overfishing and pollution. This has resulted in mass global bleaching events and a worldwide decrease of 50% in coral reef cover in the past 50 years. South Africa's coral reefs are considered high latitude marginal reefs and, as such, are good models for how corals can adapt or acclimate to environmental extremes.
We have used the high-performance computing resources provided by the CHPC to help with our work designed to protect the reefs and other vulnerable ecosystems of the South African East Coast.
We have been using high throughput sequencing and advanced molecular biological techniques to examine a number of critical questions, including:
1. Does the current Marine Protected Area network adequately protect vulnerable ecosystems from current and future threats, including climate change?
2. Can we identify cryptic species of corals and unravel coral population structure and genetic selection along the east coast of South Africa using RADSeq?
3. Can we assess and identify cryptic biodiversity in marine habitats such as coral reefs and ocean sediments using high throughput DNA sequencing and metabarcoding?
All of these methods will help us to protect these vulnerable habitats more effectively and help ensure their future in the face of threats from climate change and development.
Principal Investigator: Prof Lyudmila Moskaleva
Institution Name: University of the Free State
Active Member Count: 8
Allocation Start: 2021-01-27
Allocation End: 2021-07-26
Used Hours: 388663
Project Name: Understanding surface reactivity of solids using DFT simulations
Project Shortname: CHEM1294
Discipline Name: Chemistry
The group of Prof Lyudmila Moskaleva at the University of the Free State investigates surface reactivity of solids at the atomic level using first-principles quantum-chemical methods, molecular dynamics, statistical theory, microkinetic modelling and thermodynamics. Quantum chemical methods, in particular, those based on density functional theory (DFT), are becoming increasingly popular and powerful. They can be used successfully to achieve a mechanistic understanding at the microscopic level, which is needed to optimize functional materials (such as catalysts, electrocatalysts, semiconductors, optoelectronic materials) with respect to their target properties. We would like to mention two of our ongoing projects. One of them is a computational study on the properties of a hybrid 2D system which comprises in-plane graphene and hexagonal boron nitride (h-BN) components. We have investigated the influence of graphene or h-BN domain size on the atomic geometries, electronic properties and thermodynamic properties of prototypical graphene/h-BN hybrid systems by first-principles calculations. We have demonstrated that the band gap and the thermodynamic properties can be tuned by varying the graphene or h-BN domain size. The results of this study have been recently published in FlatChem (2020) and in C (2021). Another successful project relates to the chemistry of nanoporous gold (np-Au), a versatile material possessing interesting mechanical, optical, and catalytic properties. Our study is focused on the remarkable low-temperature catalytic activity of np-Au and its high selectivity toward partial oxidation reactions. Oxidation of methanol has been studied as a model that could be extended to higher alcohols. Our theoretical study provides a detailed mechanistic picture of a complex reaction energy network on the stepped Au(321) surface from methanol co-adsorbed with O2 all the way to methyl formate partial oxidation product. In collaboration with the experimental group of Prof Risse (FU Berlin) we studied possible scenarios for the deactivation of the catalyst. To this end we investigated several pathways for oxidation of formaldehyde to formate and further to CO and H2O. We were able to show that formate is responsible for deactivation and loss of selectivity. A relevant publication is currently in preparation. We are currently learning AIMD simulations with CP2K with the goal of exploring dynamically oxidation reactions on Au-based catalysts.
Principal Investigator: Prof Jacomine Grobler
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 12220
Project Name: Computational intelligence for supply chain optimization
Project Shortname: CSCI1170
Discipline Name: Applied and Computational Mathematics
Due to the rise of e-commerce, consumers can purchase an enormous variety of products and have it delivered to their doorstep anywhere in the world. As a result, last mile deliveries are an important competence for many logistics companies. Unmanned aerial vehicles or delivery drones, is a highly promising technology that can revolutionize the way companies do their last mile deliveries. Drones are a cleaner means of transport, do not require or use road infrastructure, and have positive implications on delivery time and cost. The efficient use of delivery drones, however, require advanced optimization algorithms to allocated deliveries to drones, and schedule these deliveries.
Researchers at the Department of Industrial Engineering at Stellenbosch University are currently working in conjunction with the University of California at Berkeley to develop a drone delivery scheduling algorithm. The development of this algorithm requires significant computational resources and here the collaboration with the Centre for High Performance Computing is critical. Many hours of computer time is required to test different algorithm variations, tune algorithm control parameters and ensure that the algorithm is robust over different data sets and changing input parameters.
Initial results indicate that up to a 30% improvement can be obtained by a delivery drone scheduling system when compared to a traditional road freight delivery service.
Principal Investigator: Dr Rhiyaad Mohamed
Institution Name: University of Cape Town
Active Member Count: 2
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 52175
Project Name: Advanced materials for oxygen evolution reaction
Project Shortname: MATS1123
Discipline Name: Chemical Engineering
Hydrogen is seen as an integral part of the sustainable energy future. Through fuel cell and electrolyser technologies, energy storage and conversion will be possible. This will resolve some of the more critical challenges faced by most renewable energy sources. Co-hosted by the University of Cape Town and operating under the Catalysis Institute, HySA Catalysis (Hydrogen South Africa Catalysis), part of a flagship programme funded by the Department of Science and Technology, aims to find efficient and market ready catalyst technologies for both hydrogen fuel cells and water electrolysers. The former has been an age-old activity with a strong internal knowledge base. On the other hand, water electrolyser research is a new venture with many unresolved and unknown challenges. The Electrolyser Research Group is a new, young and dedicated group in the area of water electrolysis within HySA Catalysis. We aim to take advantage of both experimental techniques and fundamental principles to predict and design new and better catalysts. Our research is broken down into two main categories, catalyst design and electrode membrane assembly development for commercialisation. Computational resources provided by CHPC's Lengau supercomputer enables us to perform quantum-chemical calculations to understand material properties. Whilst traditionally, unsupported iridium oxide has been used as an electrolyser catalyst, the scarcity and cost of iridium is a good justification to pursue research in supported iridium catalysts—these catalysts will of course require durable and conductive supports. Since some of these materials have never been synthesised before it is critical to methodically predict stable and electrically conductive material for use in water electrolysers. Based on chemical intuition we have identified a class of materials, doped oxides, which may possess excellent electrochemical properties. We have identified doped tin oxide and titanium dioxide as potential catalysts supports—our computational studies have now given us electronic properties of doped tin oxide. We have now identified two systems we would like to investigate further. These materials presents some interesting electronic properties. Our current efforts are now aimed at studying their stability under operating conditions.
Principal Investigator: Prof Eric K. K. Abavare
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 6
Allocation Start: 2021-01-28
Allocation End: 2021-08-18
Used Hours: 19841
Project Name: Atomic and Electronic structures of semiconductor interface systems
Project Shortname: MATS1159
Discipline Name: Physics
The Frontier Science Group (FSG) in the Department of Physics, Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, Ghana is engaged in first principles calculations as an active area of research. We investigate atomistic and electronic properties of light materials using quantum mechanical approaches.
Our researches basically focus on the study of surfaces, interface and defects in materials. All electronic devices and related components in atomic level are interfaced. We study these interfacial, defects related morphologies and how they affect the overall performance of device applications.
Experimentally, several important semiconductor materials are growth based, as a consequence complete understanding of how these growth are achieved are certainly important for device design and application.
Specifically, our interested lies with interface of silicon carbide (SiC) growth on silicon (Si) and similarly gallium nitride (GaN) on silicon for power electronic materials. These materials are important for power applications and especially crucial in hostile environment. Nevertheless, they exist in small crystal sizes and to obtain large size for device application, researchers are employing all techniques to grow them on silicon since silicon technology is well established and could be easily integrated. The key issue, which troubles the perfect layer-by-layer matching growth, is the lattice mismatch existing between the materials themselves with silicon.
By careful special crystal orientation, the materials involves above could be made to match and perfect growth achieved, which means large crystal sizes could be formed. However, the matching mechanism is not well understood and this needs computational approaches for clarification and here CHPC comes to fore in this search. The CHPC is an indispensable tool to help our group pushes the frontiers of this area for mankind in our quest for the pursuit of scientific achievement and better life through technology.
So far, our initial preliminary investigation of SiC/Si and GaN/Si interfaces are promising
Principal Investigator: Prof Evans Adei
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 9
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 1289964
Project Name: Materials For Energy and Fine Chemicals
Project Shortname: CHEM1046
Discipline Name: Chemistry
Our work at KNUST Kumasi Ghana with the support of resources from CHPC is geared towards the development of environmentally friendly catalysts for industrial applications, the mechanisms of reactions for applications in the pharmaceutical industry, and the development of functional materials that address issues of renewable energy and climate change. These are pressing real-world problems and our contributions would not have been possible without the support of CHPC.
CHPC has become an extremely important ally in our research activities. For students who are compelled to work with realistic molecular/material systems, the increase in their productivity when they start working on the CHPC platform becomes quite apparent.
The lack of funding for graduate students made the opening of modeling research experience door to undergraduate students who show interest in the work we do an obvious and pragmatic move to sustain our research activities.
Recent Research Output within the reporting period are as follows:
(i) Harun Abdullah, Richard Tia, Evans Adei (2021) Journal of Physical Organic Chemistry https://doi.org/10.1002/poc.4259
(ii) George Baffour Pipim, Richard Tia, Evans Adei (2021) Tetrahedron https://doi.org/10.1016/j.tet.2021.132306
(i) Destiny Konadu, Caroline Rosemyya Kwawu, Richard Tia, Evans Adei, and Nora Henriette de Leeuw (2021) Catalysts 2021, 11(4), 523; https://doi.org/10.3390/catal11040523
The first co-authors are all undergraduate students. The potential for future research work of some of our 2nd (Lower) students could not have been demonstrated through publication (doi.org/10.1016/j.tet.2020.131422); had it not been for the considerable generosity of CHPC CPU resources to enable us to accommodate these students.
We are aware that CHPC is catalyzing Ghana to establish its HPC center in Accra. However, with our present limited HPC resources here at KNUST Kumasi and in Ghana, we hope that CHPC will continue to support us with HPC resources in the short and medium terms.
Principal Investigator: Prof Mario Santos
Institution Name: University of Western Cape
Active Member Count: 17
Allocation Start: 2021-01-29
Allocation End: 2021-08-12
Used Hours: 1161691
Project Name: Cosmology with Radio Telescopes
Project Shortname: ASTR0945
Discipline Name: Astrophysics
The Centre for Radio Cosmology (CRC) at UWC aims to fully explore the use of the next generation of radio telescopes for measurements in cosmology, in particular with MeerKAT and SKA. CRC staff is closely involved in the South African and international SKA projects. About 10 postdoctoral researchers and 10 PhD/MSc students are doing their research within the CRC. Our work at the CHPC is mostly focused on running simulations of signals we expect to observe with MeerKAT/SKA and test how well radio telescopes can constrain certain cosmological models. This is a crucial component in the process to use these telescopes for cosmology since we need to understand the signals we are observing. The project include simulations of the hydrogen emission from the early universe, simulations of the contaminants, simulations of the telescope itself and analysis of the constraining power of such telescopes on cosmological models.
Principal Investigator: Prof Timothy Egan
Institution Name: University of Cape Town
Active Member Count: 3
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 102890
Project Name: Haem chemistry and interactions with antimalarial compounds
Project Shortname: CHEM0935
Discipline Name: Chemistry
The Bioinorganic Research Group at the University of Cape Town is led by Professor Timothy J. Egan. The work conducted by this group largely focuses on malaria, with this this programme aimed at understanding the interaction of new compounds with the malarial pigment haemozoin. Haemozoin is a crystal formed in the parasite from haem, the iron containing group in haemoglobin that is responsible for binding oxygen. The parasite lives in host red blood cells, where it digests large quantities of haemoglobin, releasing toxic free haem. Haemozoin formation is a process by which the parasite rids itself of this toxin. Certain antimalarial drugs act by inhibiting haemozoin formation. This is thought to occur by binding to the growing haemozoin crystal. A systematic process of identifying compounds available commercially from ChemDiv which has large libraries of drug-like molecules has been used to narrow down 25,000 compounds for docking with the surface of haemozoin. These have been docked and the top-ranked compounds identified, filtered for toxicity and other undesirable properties. A short list of about 20 compounds is currently being compiled for purchase and testing for inhibition of synthetic haemozoin formation and activity against malaria parasites grown in laboratory culture.
In addition we have also explored the effects on binding to of pyridine-to-pyrimidine ring interconversion and CF3 defluorination on PI3K, and important cell signalling protein important in cancer and potentially useful in malaria treatment. This latter work is currently under review for the Journal of Physical Chemistry. The work should aid in computation drug design.
Principal Investigator: Prof Gerrit Basson
Institution Name: Stellenbosch University
Active Member Count: 6
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 614636
Project Name: CFD modelling of Hydraulic Structures in Rivers
Project Shortname: MECH0985
Discipline Name: Computational Mechanics
The Institute for Water & Environmental Engineering Stellenbosch University has 3 Doctoral students and 3 Post-Doctoral student who are investigating CFD modelling of hydraulic structures with complex vortices and sedimentation. Key aspects of the research are related to the improved prediction of bridge pier scour, to investigate the addition of flares on spillways, to reduce rock bed scour in plunge pools at spillways, to optimize the design of vortex settling basins, sand traps and settlers for sediment removal at hydropower and water abstraction plants. The simulations are too computationally intensive for standard computers and the CHPC assists in the research by allowing the students to more extensively explore the hydraulic behaviour of these structures.
Principal Investigator: Dr Ndanduleni Lethole
Institution Name: University of Fort Hare
Active Member Count: 4
Allocation Start: 2021-01-29
Allocation End: 2021-07-28
Used Hours: 104426
Project Name: Computational Studies of MxPt1-x (M; Mn, Fe, Co and Ni) alloys for magnetic data storage and biomedical applications.
Project Shortname: MATS1309
Discipline Name: Material Science
The research group is located under the department of Physics, University of Fort Hare. The group started in February 2020 and is composed of the Principal Investigator and two BSc Honours students. Currently, the group entirely depends on the Centre of High Performance Computing (CHPC) facility since it uses Material Studio modeling and simulation environment accessed via the CHPC as its only simulation package. The group aims to grow to at least two MSc and 2 Honours students in the year 2022 and also acquire a license for the Vienna ab initio simulation package. The group is currently undertaking two computer simulation studies, namely; first-principles calculations of M-Pt (M: Mn, Fe, Co and Ni) alloys for potential application as advanced performance permanent magnets, new storage of ultra-high-density magnetic data and in biomedical applications; and Zn/CaMn2O4 for potential application in Zn and Ca ions rechargeable batteries. The M-Pt alloys is significant since there has been a remarkable increase in data storage density of magnetic disk drives in the past few decades which is projected to be restricted by the "super-paramagnetic limit" (loss of data due to activated fluctuations of magnetization) in the near future. Super-paramagnetic limit can be overcome by designing new magnetic storage materials which are energetical, electronical, magnetical, mechanical and dynamical stable. Thanks to modern simulation packages and sufficient computing resources available at the CHPC, augmentation of various properties of the crystal structure is possible by theoretically calculating the forces acting on the nuclei. Simulations are set up on the local desktop computer using the BIOVIA Materials Studio modeling and simulation environment and submitted to the CHPC computing resources using the CASTEP code. On the negative note, the progress of the project has been disadvantaged by the ongoing lockdown in South Africa. However, few results have been attained. These include; heats of formation, elastic constants, density of states and band structure on the MnPt alloy.
Principal Investigator: Prof Shazrene Mohamed
Institution Name: South African Astronomical Observatory
Active Member Count: 4
Allocation Start: 2021-01-28
Allocation End: 2021-07-27
Used Hours: 600997
Project Name: 3D rad-hydro simulations of the outflows of evolved stars
Project Shortname: ASTR0880
Discipline Name: Astrophysics
The computational stellar astrophysics group (PI. S. Mohamed) is currently based at the South African Astronomical Observatory and the University of Cape Town. The research uses supercomputers, powerful computational machines at the CHPC, to simulate the complex physics involved in the interactions of stars in the final phases of evolution. The interactions include exchanges with a nearby companion star or planet that can lead to the formation of spirals, disks, jets and even violent stellar explosions. By comparing the models with observations from powerful telescopes, these simulations help us understand how solar systems evolve and ultimately, the fate of our own star, the sun.
Principal Investigator: Dr Nkululeko Emmanuel Damoyi
Institution Name: Mangosuthu University of Technology
Active Member Count: 3
Allocation Start: 2021-01-29
Allocation End: 2021-07-28
Used Hours: 47959
Project Name: Surface Reaction Mechanisms
Project Shortname: CHEM1105
Discipline Name: Chemistry
My present area of research is in Computational Chemistry (CC). Computer technological advances make it possible to do research in CC in order to model chemical reactions and predict chemical reaction properties which are difficult or impossible to determine from normal laboratory experiments. Most of my present research involves utilising CC to predict the likely chemical reaction mechanisms that produce valuable organic compounds from alkanes with use of catalysts and also without catalysts.
Principal Investigator: Prof Josua Meyer
Institution Name: University of Pretoria
Active Member Count: 6
Allocation Start: 2021-01-29
Allocation End: 2021-07-28
Used Hours: 657332
Project Name: Fundamentals of forced and mixed convection in heat exchangers
Project Shortname: MECH1094
Discipline Name: Computational Mechanics
I am leading the Clean Energy Research Group (CERG) at the University of Pretoria. We focus on thermal solar heat exchangers, analyzing and optimizing the heat transfer rates. We investigate the use of single phase and multi-phase fluids (and newer environmentally friendly working fluids and gasses) and enhanced surfaces, including surfaces with Nano-layer that enhances surfaces and heat transfer rates. Other applications that are being studied are the cooling of microelectronic components and gas turbine blades. Over the past few years we have collaborated and/or exchange students with MIT, Imperial College London, Edinburg University, University of Lagos and Mauritius University. Our students have access to the HPC from these universities and in all cases they could complete their numerical work faster than using the available numerical capabilities at our partner universities.
Principal Investigator: Dr Bubacarr Bah
Institution Name: African Institute for Mathematical Sciences
Active Member Count: 12
Allocation Start: 2021-02-01
Allocation End: 2021-08-12
Used Hours: 41322
Project Name: Large Scale Computations in Data Science
Project Shortname: CSCI0972
Discipline Name: Applied and Computational Mathematics
This research programme is the programme of the African Institute for Mathematical Sciences (AIMS) South Africa Data Science Research Group. The group conducts research in the theory, computation and applications of deep learning (machine learning with deep neural networks). The 5 current projects running are as follows. 1) Verifying robustness of neural networks. This will address the frequent failure of deep learning models. 2) Application of Graph Convolutional Neural Networks (GCNNs) to Computer vision problems. GCNNs were originally designed for graph data as input. However, this project seeks to understand how well will GCNNs perform compared to standard CNNs for computer vision tasks. 3) Deep Learning for Group Testing applied to COVID-19. Instead of designing deterministic group testing settings, the projects aim to learn optimal settings for group testing using data. 4) Automatic animal identification using deep learning. This is an application in ecology where we want to automatically identify animals without having to tag them. 5) Post-processing the output of image segmentation deep learning algorithms in a bit to improve their results. The student is working with images of fruit trees and hence one of the outcome would be to improve projections on yields. All projects are progress well. Project 2 will be concluded soon with the student planing to graduate in December 2021.
Principal Investigator: Prof Nelishia Pillay
Institution Name: University of Pretoria
Active Member Count: 12
Allocation Start: 2021-02-01
Allocation End: 2021-07-31
Used Hours: 327111
Project Name: Nature Inspired Computing Optimization
Project Shortname: CSCI0806
Discipline Name: Computer Science
The NICOG (Nature Inspired Computing Optimization Group), based at the University of Pretoria, employs machine learning and optimization techniques, taking an analogy from nature, to solve real world problems to attain the sustainable development goals defined by the United Nations. This research is solving problems in the health sector, e.g. disease prediction, agriculture, e.g. disease, crop modelling and streamflow prediction and industry, e.g. logistics, scheduling and classification problems. As we move into the fourth industrial revolution, machine learning is playing a major role in solving these problems. The algorithms used to solve these problems are computationally expensive and hence high performance computing is needed for the implementation of the algorithms. We are at point where we need to make these approaches easily accessible to non-experts. Hence, one of the current foci of NICOG is the automated design of machine learning techniques to produce off-the-shelf tools for non-experts. Such automated design, which is essentially an optimization problem, is also computationally expensive and thus also requires high performance computing.
Principal Investigator: Prof Rebecca Garland
Institution Name: University of Pretoria
Active Member Count: 18
Allocation Start: 2021-02-01
Allocation End: 2021-07-31
Used Hours: 220188
Project Name: Development of the first African-based earth system model VRESM and its projections of future climate change over Africa
Project Shortname: ERTH0859
Discipline Name: Earth Sciences
The first African-based Earth System Model (ESM) is under development at the Council for Scientific and Industrial Research (CSIR) in South Africa, through collaboration with the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia and South Africa's Centre for High Performance Computing (CHPC). The new ESM is unique in the sense that it is being developed in Africa and through the lens of African and Southern Hemisphere climate processes. That is, the reliable simulation of climate issues critical to Africa, such as the occurrence of drought in in response to El Niño events, wide-spread flooding over Mozambique due to landfalling tropical cyclones and increased occurrences of heat-waves under global warming are key issues in the development of the new model. The model development process is also focussed on the Southern Ocean's role in regulating southern African climate, as well as the global climate through the absorption of carbon dioxide in the ocean. The development and application of an Earth System Model is a computationally extensive endeavour, and was impossible to undertake in South Africa before the Lengau cluster of the CHPC became available in 2018. The CSIR-CHPC partnership towards development of the new model has in 2018 led to the most detailed set of projections of future climate change over southern Africa obtained to date. These simulations have provided new insight into the plausible impacts of climate change in South Africa, including how the frequencies of droughts, heat-waves and landfalling tropical cyclones may change over the next few decades. These projections were key in informing South Africa's Third National Communication on Climate Change, which was published in 2018 under the science leadership of the CSIR. Moreover, the research papers based on the projections have directly informed the Special Report on Global Warming of 1.5 ºC of the Intergovernmental Panel on Climate Change (IPCC).
Principal Investigator: Prof Regina Maphanga
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 7
Allocation Start: 2021-02-02
Allocation End: 2021-08-18
Used Hours: 226662
Project Name: Energy Storage Materials
Project Shortname: MATS0919
Discipline Name: Material Science
The Design and Optimisation group is a group under Operational Intelligence Impact Area in the Next Generation Enterprises and Institutions cluster of CSIR and uses multiscale methods to develop novel energy storage materials with desired properties and enhanced properties for existing materials. The envisaged outcome of this project is to develop energy storage materials models with improved performance and consistency, which will be cheaper and environmentally benign. The ever-increasing global energy needs and depleting fossil-fuel resources demand the pursuit of sustainable energy alternatives, including renewable sources to replace carbon intensive energy sources. Sustainable and renewable energy is considered to be the most effective way to minimize CO2 emissions. Hence, finding sustainable energy storage technologies is vital for optimally harnessing the renewable energy.
Computational methodologies are very effective in predicting material-structure-property correlations. In this project simulation methods and models based on parallel computing are developed to probe energy materials properties. Centre for High Performance Computing (CHPC) resources are used to develop these parallel computing methods and algorithms for large and complex material models. Thus, CHPC resources provide a platform to simulate the evolution of material properties at a wide range of external conditions that are not accessible experimentally and fast track acquisition of results. The major of aim of this project is to create a predictive, reliable and robust set of models of materials for energy storage across the materials modelling length scale, leading to integrated multi-scale capability.
Principal Investigator: Prof Daniel Joubert
Institution Name: University of the Witwatersrand
Active Member Count: 18
Allocation Start: 2021-02-02
Allocation End: 2021-08-18
Used Hours: 1080522
Project Name: Energy Materials: Numerical explorations
Project Shortname: MATS0800
Discipline Name: Physics
Imagine designing a material atom by atom. Imagine analysing the properties of tomorrow's novel materials before they exist. Computational materials science plays an important role at every of level of the design and engineering of new materials. The rapid advances in computer processing power and memory has given virtual, fundamental, materials design a boost. The first problem faced in designing a virtual material, a material that has not yet been made, is to find the stable configurations in which the atoms in the mix will settle into. The designer must find the configuration of electrons and nuclei, the building blocks of atoms, which result from the interactions among a number of particles that, even for a small piece of material, exceeds the number of particles of sand on all the beaches in the world. This daunting task has a simple solution. Instead of finding the configuration of the real material, the problem is solved for a fictitious material where the constituent particles do not interact. All that is necessary is to find the particle density distribution of the fictitious material, which by design, is the same as that of the real material. A Noble prize was awarded for this idea to the chemists Walter Kohn and John Pople in 1998.
A group of postgraduate students at the University of the Witwatersrand use the computer power at the national Centre for High Performance Computing to conduct virtual experiments on existing and novel materials to examine their potential as energy harvesters. They examine the potential of a selection of materials that can be used as cheap components in solar cells, materials that can generate electricity from waste heat and materials that can be used to split water molecules to extract hydrogen for energy production.
Principal Investigator: Dr Michael Owen
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-02-02
Allocation End: 2021-08-01
Used Hours: 155867
Project Name: Industrial heat exchanger simulation and analysis
Project Shortname: MECH1210
Discipline Name: Computational Mechanics
The Thermofluids Division at Stellenbosch University's Department of Mechanical and Mechatronic Engineering has a long history of research in industrial heat exchangers and has conducted pioneering work in the field of dry cooling for large thermal power plants in particular.
Research in this field continues with a focus on understanding and improving the performance of these water conserving cooling systems under a variety of ambient conditions (including strong winds). Work of this nature benefits greatly from the CHPC which allows us to investigate an unprecedented number of scenarios in a short amount of time using computational fluid dynamics (CFD) and parallel computing.
Dry cooling is a key technology for South Africa's energy sector since thermal energy production (coal, nuclear, combined cycle gas and solar thermal) is typically a water intensive process which South Africa can ill afford considering the arid nature of our country. Ensuring robust operation of dry cooling systems can contribute significantly to the sustainability of South Africa's energy production.
Principal Investigator: Prof Raymond Hewer
Institution Name: University of KwaZulu-Natal
Active Member Count: 8
Allocation Start: 2021-02-02
Allocation End: 2021-08-01
Used Hours: 17507
Project Name: The design and discovery of small-molecule inhibitors of HIV-1 integrase activity
Project Shortname: CHEM0877
Discipline Name: Chemistry
Our research group, led by Dr Raymond Hewer, within the Discipline of Biochemistry at the University of KwaZulu Natal seeks to design and discover novel compounds with therapeutic potential against HIV-1 and Alzheimer's Disease; two diseases that represent a high burden of disease in South Africa. Our primary approach to identify new compounds involves the use of state-of-art molecular modeling programs which we access through agreement with the CHPC. A number of compounds identified in this manner have proven to be effective in vitro with findings published in various Journal articles
Principal Investigator: Prof Peter Teske
Institution Name: University of Johannesburg
Active Member Count: 3
Allocation Start: 2021-02-03
Allocation End: 2021-08-02
Used Hours: 208377
Project Name: Coevolution of mangroves and mangrove-associated crabs
Project Shortname: CBBI0979
Discipline Name: Bioinformatics
Who?
Centre for Ecological Genomics and Wildlife Conservation, University of Johannesburg
What?
Analysis of DNA and RNA data from a large portion of animal genomes. A noteworthy development in 2021 is the publication of a paper in Science Advances (currently in press) that uses genomic data to explain why the KwaZulu-Natal sardine run happens. This is the highest-ranking journal ever to be published at our department (impact factor: 14.1).
Why?
To improve the management of commercially exploited species, optimise conservation efforts aimed at saving endangered species from extinction, and understand evolutionary relationships in nature.
How?
Genomic data sets generated using massive parallel sequencing technology results can only be assembled by means of supercomputers. Using CHPC, we are able to assemble a huge number of short reads into a meaningful sequence that reflects their order in the animal genome.
Progress?
CHPC and the valuable expertise it has gathered in one place has allowed our small lab to continue punching well above its weight. Despite the challenges related to the lockdown, 2021 promises to be an even more productive year than 2020. This would not have been possible without the CHPC.
Principal Investigator: Prof Ozlem Tastan Bishop
Institution Name: Rhodes University
Active Member Count: 19
Allocation Start: 2021-02-03
Allocation End: 2021-08-02
Used Hours: 7359481
Project Name: Structural Bioinformatics for Drug Discovery
Project Shortname: CBBI0867
Discipline Name: Bioinformatics
Research Unit in Bioinformatics (RUBi) at Rhodes University is specialized in structural bioinformatics research. Structural bioinformatics focuses on the identification and analysis of structure, dynamics and interaction of biological macromolecules in 3D space. While doing so, it aims to make connections between sequence – structure and function of macromolecules. Information on structure is particularly important as you can derive enormous amounts of information.
RUBi's main research interest is early drug discovery. Computer aided drug design technology has greatly revolutionised the drug research and development process. Traditional computational drug discovery approach includes large scale structure based and/or ligand based in silico docking combined with molecular dynamics and binding free energy calculations. All these computationally expensive calculations require to use HPC facilities.
In the last few years, RUBi has been developing some novel protocols by merging computational chemistry, structural bioinformatics and biophysics approaches and combining with traditional methods. These combined approaches also require HPC facilities. We expect that our protocols will advance our and others research. We have already been testing the combined approaches in different cases and a number of hit compounds identified; including identification of South African natural compounds as hits to cancer, malaria, trypanosoma and HIV.
Principal Investigator: Dr Samuel Atarah
Institution Name: University of Ghana
Active Member Count: 1
Allocation Start: 2021-02-04
Allocation End: 2021-08-26
Used Hours: 143240
Project Name: Ab initio studies of electronic and magnetic properties of selected elements
Project Shortname: MATS1162
Discipline Name: Physics
I am Samuel Atarah, Department of Physics, University of Ghana. My group is part of the larger Condensed Matter Physics group. Heavy elements continue to be an interesting area of research. This is because of their relevance in radioactive applications and in particular the search for the 'Island of Stability'. In that Island a number of very heavy elements are predicted to be very stable. When they are discovered without much knowledge on their physico-electronic properties, ab initio studies on them are a good lead to their full (experimental) characterization such. Equally important especially for future clean energy generation are the perovskites. A lot is desired to be understood of electronic and optial properties of different possible combinations of elements in a perovskite-like structures. Computational studies are crucial in this type research. We mainly use the Quantum Espresso suite to study electronic, optical and phonon properties of material which applies Density Functional Theory. With computational resources at Lengau CHPC, we are able to engage in the above areas of investigation to hopefully contribute to contemporary research results.
Principal Investigator: Prof Tien-Chien Jen
Institution Name: University of Johannesburg
Active Member Count: 22
Allocation Start: 2021-02-05
Allocation End: 2021-08-18
Used Hours: 515864
Project Name: Developing an Innovative Metal Matrix Composite Membrane for Hydrogen Purification
Project Shortname: MATS0991
Discipline Name: Material Science
Hydrogen is an excellent energy carrier because of its high enthalpy of combustion with less environmental impact. Hydrogen fuel can be directly used in cars, ships, aircraft, and other transportation equipment. Although there are many problems to be resolved for implementation of hydrogen technology, it is generally believed that hydrogen will become a key energy carrier in this century. Until sufficient quantities of hydrogen can be derived from renewable energy sources, it will have to come from hydrogen separation using existing fuels (e.g., diesel, natural gas) as the feedstock. Hydrogen separation units can generate highly purified hydrogen from the feedstock gas mixtures, however, a certain amount of CO and other chemicals will make it into the hydrogen stream and cause catalyst failure on the fuel cell electrode. This significantly deteriorates the activity of the catalyst layer, particularly for Proton-Exchange Membrane (PEM) fuel cells. This deterioration greatly reduces the efficiency of power generation and the life cycle of the fuel cell. Overcoming this obstacle is necessary for the wider use of highly efficient fuel cells. The objective of this mproposal is to develop a new composite metal membrane (CMM) that can achieve extremely high flux and selectivity of the hydrogen separation and achieve the lifetime goal set by the U.S. Department of Energy (DOE). To accomplish these goals, a research team formed by the University of Johannesburg (UJ), Council for Scientific and Industrial Research (CSIR) and Hydrogen South Africa (HySA) will combine two major technologies: Electrostatic-Force-Assisted Cold Gas Dynamic Spraying (ECGDS) and Laser-Assisted Manufacturing (LAM) to develop this new type of membrane. The proposed new membrane will eliminate the hydrogen embrittlement, catalyst failure, and avoid the difficulty of forming a defect-free, ultrathin Pd film on the substrate surface. The proposed CMM will exhibit superior reliability and can be produced with a relatively low-cost and scalable process. Along with the development process, numerous numerical methods are utilized to simulate the material properties and behaviour as well as the optimization fabrication procedure. To accomplish these task HPC services from CHPC plays a crucial role in understanding the numerous unknown phenomena's. This service enables the research to be simulated from feature scale to reactor scale .Note that this project is currently supported by the NRF for 2017-2021
Principal Investigator: Dr Nikhil Agrawal
Institution Name: University of KwaZulu-Natal
Active Member Count: 2
Allocation Start: 2021-02-05
Allocation End: 2021-08-04
Used Hours: 129874
Project Name: Molecular Modeling of proteins using different computational tools
Project Shortname: HEAL1189
Discipline Name: Health Sciences
We work on proteins that are involved in diseases such as HIV, Alzheimers, Parkinson's etc. Millions of people around the world including South Africa are suffering from these diseases. In our research, We are utilizing molecular modelling techniques such as molecular dynamics simulations to investigate their binding with drug molecules, membrane and material surfaces. The outcome of our research will be helpful to design new or improved candidate drug molecules related to these diseases and will also further enhanced our knowledge on these disease-causing proteins.
Principal Investigator: Prof Kenneth Ozoemena
Institution Name: University of the Witwatersrand
Active Member Count: 2
Allocation Start: 2021-02-08
Allocation End: 2023-05-18
Used Hours: 2832
Project Name: Materials Electrochemistry and Energy Technologies
Project Shortname: MATS1353
Discipline Name: Chemistry
Our research group is called the "Materials Electrochemistry and Energy Technologies" (aka MEET). We are at the School of Chemistry of the University of the Witwatersrand (Wits). Our work is part of the NRF-funded South African Research Chair Initiative (SARChI).
Our research is focused on developing clean and affordable energy (such as lithium-ion batteries and fuel cells. We also study how to develop devices that can detect diseases such tuberculosis, cholera, and cervical cancer. For more information, please check our website: www.ozoemenagroup.com.
CHPC helps our research a lot as it allows us to first understand how the materials we make in the laboratory work before we even start doing our experiments in the laboratory.
Our projects are going well, Masters and Doctoral students are being graduated, and we hope that our work with the CHPC will continue to support us.
Principal Investigator: Dr Phumlani Mdluli
Institution Name: Mintek
Active Member Count: 2
Allocation Start: 2021-02-08
Allocation End: 2022-04-28
Used Hours: 17483
Project Name: Gold SERS nanoprobes as sensors for detection of trace levels of acidic pharmaceuticals in water
Project Shortname: MATS0891
Discipline Name: Material Science
Mintek has been supported by CHPC for all projects on molecular modelling. Our work focusses mainly on the simulation of metal catalysts focusing mainly on understanding catalytic mechanism for adsorption of carbon monoxide. In this project, we intend to on the surface of metal nanoparticles supported with graphite. We have made some strides to understand these mechanism. We would like to request the use of CHPC facilities to further simulate the magnetic properties of graphite supported Cobalt catalysts which is hinted as the driving force during for the catalytic conversion of CO into useful gases .
Principal Investigator: Prof Oleg Smirnov
Institution Name: Rhodes University
Active Member Count: 8
Allocation Start: 2021-02-09
Allocation End: 2021-08-08
Used Hours: 5494
Project Name: MeerKAT Imaging
Project Shortname: ASTR0967
Discipline Name: Astrophysics
The RATT/RARG group has been using CHPC as a testbed for distributed radio astronomy data processing applications. Future telescopes such as the SKA will required massively distributed implementations for their data processing. RATT/RARG has been using tools such as Dask and Numba to develop these, and to test them at CHPC.
Principal Investigator: Prof Thulani Makhalanyane
Institution Name: Stellenbosch University
Active Member Count: 10
Allocation Start: 2021-02-09
Allocation End: 2021-08-08
Used Hours: 43543
Project Name: Metagenomics of extreme environments
Project Shortname: CBBI1023
Discipline Name: Bioinformatics
Our group aims to understand the contribution of microbial communities in the environments. The challenge is that such microbiomes are numerically abundant and may include billions of microorganisms. We apply next generation sequencing tools to identify the specific microbiomes and characterize their ecosystem contributions. Our projects include studies on human gut, terrestrial and marine microbiomes. For example, the study by Kabwe et al who graduated PhD studies this year, used metagenomics to assess fungi found in gut microbiomes. This first study on the gut microbiomes from urban and rural South African locations also used these sequence data analysed using CHPC resources to assess antibiotic resistance genes in individuals.
Principal Investigator: Prof Nithaya Chetty
Institution Name: University of Pretoria
Active Member Count: 5
Allocation Start: 2021-02-09
Allocation End: 2021-08-12
Used Hours: 354543
Project Name: Advanced Computational Resources for members of ASESMA
Project Shortname: MATS1050
Discipline Name: Material Science
Work on two dimensional (2D) materials is now firmly established as an area of excellence at the theoretical and computational solid state physics group of the University of Pretoria. The focus in the past years was mainly on the study of the popular 2D materials such as molybdenum disulphide, graphene and boronitrene. The 2D hexagonal-like transition metal oxides (TMO) materials such as molybdenum dioxide (MoO2) monolayer is also predicted to be thermodynamically stable, however, a detailed study on the structural, electronic and dynamical properties of these new oxide materials is still missing. Therefore, research about these properties is important and could bring them from the laboratory to commercial products. Recently (over the past two years) a 2D bilayer of zeolite has been experimentally synthesized by scientists in the USA. This is an area of promising new research for computational studies, and worth to be explored. We are mainly interested in the use of these 2D materials for catalysis such as CO catalytic oxidation. To achieve these goals, creating complex systems such as introducing defects or adatoms will enhance the properties of the MoO2 and the 2D zeolite model. To ensure that the introduced defects are isolated, the study should be done on a large supercell which is computationally expensive, and could not be done on a local computer. Powerful computers with large and faster CPUs such as CHPC cluster are necessary. Our results provide detailed understanding and important information to the 2D materials experimental and computational community.
Principal Investigator: Prof Paulette Bloomer
Institution Name: University of Pretoria
Active Member Count: 7
Allocation Start: 2021-02-10
Allocation End: 2021-08-09
Used Hours: 528892
Project Name: Molecular Ecology and Evolution Programme
Project Shortname: CBBI1030
Discipline Name: Bioinformatics
The Molecular Ecology and Evolution Programme (MEEP) is a research group in the Division of Genetics, Department of Biochemistry, Genetics and Microbiology at the University of Pretoria. We investigate the evolutionary history and current genetic diversity of many vertebrate species (marine and terrestrial), using genetics and genomics tools, and try to link the patterns we see with ecological and human-mediated events. We currently have two projects (Cape buffalo and bowhead whales) that make use of the CHPC resources, with additional projects in the pipeline.
Genomics of Cape Buffalo is a hot topic and is of interest to the general public and wildlife ranchers in particular. Cape buffalo are bred on private game ranchers for a one or a few economically important traits, such as horn length and body size. The effects of this artificial selection, as well as breeding South African buffalo with East African buffalo, is unknown. We can use genomics to evaluate the risks and benefits of these practices for the short and long term (evolutionary) conservation prospects of the species
The genetic work on the bowhead whale uses empirical and simulation data to investigate the effects that different Nearctic human cultures exploiting the whale had on the abundance of the population. This means that genetic data may have more resolution than previously thought and is sufficient to detect subtle changes in population size in relation to exploitation and climate change. This is of great importance as there are many applications. We believe that the work will have sufficient impact to be of interest to the general public in South Africa and internationally.
The development of a new molecular clock calibration method for tropical reef species makes use of simulations for validation. This calibration enables genetic time to be related to chronological time. This method may then be applied to empirical data to identify driving factors for increases in population size of reef dependent species in the tropics.
Principal Investigator: Prof Liliana Mammino
Institution Name: University of Venda
Active Member Count: 3
Allocation Start: 2021-02-10
Allocation End: 2021-08-09
Used Hours: 849680
Project Name: computational study of biologically active molecules of natural origin
Project Shortname: CHEM0959
Discipline Name: Chemistry
WHO?
The group consists of one professor, one student who has just completed his M.Sc. project and is starting his PhD one, and one second-year PhD student, at the University of Venda.
WHAT?
We focus on the study of biologically active molecules, that is, molecules that can have an effect on human organisms and can be potentially interesting for drug development. We are currently studying molecules with antidiabetic, anticancer, antimalarial and antiviral properties. Some molecules of the largest family that we have been studying (acylphloroglucinols) are currently being evaluated for possible anti-covid activity, what contributes to recommend the continuation and deepening of our studies.
WHY?
We study the properties of the selected molecules.
The pharmaceutical actions of molecules depend on their properties. Therefore, the results from our studies provide information that is important for specialists who work on further steps of drug development. These specialists can use the information from our studies to select the most suitable molecules and predict the activities of modified molecules (where the modifications are meant to improve the activity).
Drug development is not a fast process. It requires a lot of contributions from different specialists. Our contribution consists of information about the molecular properties of promising molecules.
HOW?
We use calculations to find the properties of the molecules that we consider.
There are different approaches for these calculations. The ones that give more reliable results require huge time to complete. Without the use of the CHPC, it would not have been possible to obtain most of the results that we have obtained. That is why using the CHPC is essential for us to be able to conduct our research.
HOW IS THE PROJECT PROGRESSING?
The project is progressing satisfactorily, despite some challenges caused by the pandemic.
Principal Investigator: Dr Sarah Blyth
Institution Name: University of Cape Town
Active Member Count: 7
Allocation Start: 2021-02-10
Allocation End: 2021-08-26
Used Hours: 21729
Project Name: LADUMA: Looking At the Distant Universe with the MeerKAT Array
Project Shortname: ASTR1354
Discipline Name: Astrophysics
The LADUMA (Looking At the Distant Universe with the MeerKAT Array) survey is one of the approved Large Survey Projects to be observed on the MeerKAT radio telescope. The team is co-led by Sarah Blyth (University of Cape Town), Benne Holwerda (University of Louisville), and Andrew Baker (Rutgers, The State University of New Jersey) and comprises more than 80 South African and international scientists. The aim of the survey is to observe the neutral hydrogen gas in galaxies over cosmic time to understand how galaxies have evolved since the universe was only one-third of its current age. LADUMA will be the deepest survey of neutral hydrogen emission to date and will enable direct detections of neutral hydrogen well beyond the local universe for the first time. The survey will observe the sky for thousands of hours, generating an enormous radio dataset that will need to be calibrated and turned into 3D data cubes (2D spatial images x 1D frequency information). Achieving this goal will require powerful computers that can process the data in reasonable timeframes. The CHPC computers are excellent for the kinds of processing required for LADUMA, and the survey team is using its early data and the CHPC computers to test calibration and imaging software and determine the ideal strategy for processing future, much larger datasets. This work is vital to the success of the survey overall and is progressing well, en route to a first science data release during 2021.
Principal Investigator: Prof Gill Enslin
Institution Name: Tshwane University of Technology
Active Member Count: 2
Allocation Start: 2021-02-11
Allocation End: 2021-08-10
Used Hours: 26722
Project Name: Enzyme and receptor targets for phytomedicines, molecular modelling and dynamic simulation of ligand-target interactions
Project Shortname: HEAL1289
Discipline Name: Health Sciences
One of the projects of the Phytomedicines research group, led Professor Alvaro Viljoen at the Tshwane University of Technology, is the investigation and validation of the use of traditional herbal medicines for their anti-inflammatory, anti-epileptic and anxiolytic effects. In some patients, currently used anti-epileptic medicines are not fully effective in controlling seizures or have adverse effects that impinge on the quality of life of the patient. In South Africa a large number of our population rely on traditional medicines, and do not seek allopathic medicine as their first point of health care advice. These herbal medicines require scientific validation to contribute to safety, quality and efficacy in use. Plants also offer a wealth of new molecules for study, and the potential to find newer and better treatments for many ailments, including central nervous system disorders.
Extracts from plants, identified from the literature for their use as anxiolytics or anticonvulsants, were tested for bioactivity in an animal model, the zebrafish larva. Potentially active compounds have been isolated identified and their chemical structures determined. These compounds are being subjected to virtual screening and further investigation using computational chemistry to identify the targets in the central nervous system and to simulate molecular interactions between the molecules and the biological target. This final step relies totally on the computing power of the CHPC platform
Principal Investigator: Mr Johan Bester
Institution Name: Sebenzana Consulting
Active Member Count: 2
Allocation Start: 2021-02-12
Allocation End: 2023-02-09
Used Hours: 11494
Project Name: Sebenzana Consulting CFD
Project Shortname: INDY1393
Discipline Name: Computational Mechanics
Sebenzana is an engineering consultancy specializing in asset performance and management. The majority of assets are related to power generation, however there is some exposure in other industrial processes as well. The CHPC is used mainly for CFD models, greatly reducing investment costs in terms of hardware but also a significant reduction in the project timeline. The CFD simulations are typically so complex that without the CHPC, the investigation won't be financially viable for a client.
Principal Investigator: Prof Cedric McCleland
Institution Name: Nelson Mandela Metropolitan University
Active Member Count: 1
Allocation Start: 2021-02-15
Allocation End: 2021-08-14
Used Hours: 23578
Project Name: Reactive Intermediates
Project Shortname: CHEM1394
Discipline Name: Chemistry
Reactive intermediates like radicals, cations and radical-cations are key to numerous chemical reactions. The purpose of this study is to gain a better theoretical understanding of these species, especially radical-cations, so as to utilise them more effectively in chemical reactions.
Principal Investigator: Dr Zibo Keolopile
Institution Name: University of Botswana
Active Member Count: 2
Allocation Start: 2021-02-15
Allocation End: 2021-09-09
Used Hours: 375773
Project Name: Physics
Project Shortname: MATS1303
Discipline Name: Physics
The theoretical and computational group is based on the Department of Physics, University of Botswana. The main researchers are Dr Z. G. Keolopile and Mr L Radisigo and our main research is focused on but not limited to the following streams;
Electronic Structure Methods: Development of electronic structure atomic-orbital-based methods for molecular and crystalline systems. Development of combinatorial-computational approaches based on quantum chemical engines. Suppressing barriers on potential energy surfaces obtained with electronic structure methods. Electron binding energies in molecular anions. Electronegativity of molecular building blocks.
Computational Chemistry of Materials and Interfaces: Materials for hydrogen storage. Design of clathrate materials.
Biophysics: Intermolecular proton transfer induced by excess charge. A typical tautomers of charged nucleic acid bases.
Intermolecular Interactions: Hydrogen bonding. Solvation of ions. Many-body interactions. Symmetry rules for valence repulsion energy. Partial wave expansion and damping phenomenon for the dispersion energy. Basis set consistency in calculations of intermolecular interaction energies.
Chemical Dynamics: Solvent assisted proton transfer. Dynamics on deformed potential energy surfaces.
Our electronic structure calculations are very much dependent on solving the Schrodinger equation. These calculations cannot be solved by hand therefore these studies require the use of high performance computers (HPC). The group has progressed well using resources from CHPC and a number of manuscripts are under preparation.
Principal Investigator: Dr John Mack
Institution Name: Rhodes University
Active Member Count: 2
Allocation Start: 2021-02-15
Allocation End: 2021-08-14
Used Hours: 29056
Project Name: Rational design of phthalocyanine, porphyrin and BODIPY dyes
Project Shortname: CHEM0796
Discipline Name: Chemistry
The Institute for Nanotechnology Innovation (INI) at Rhodes University under director Prof. Tebello Nyokong carries out research related to the use of molecular dyes in biomedical applications, such as cancer treatment through photodynamic therapy, the use of antimicrobial photodynamic chemotherapy for treating hospital superbugs and as sensors for ions that are harmful to human health.
Other applications of interest include wastewater treatment and the development of optical limiters to protect human vision,such as in aviation safety in protecting pilots from the irresponsible use of laser pointers during runway approaches. It is important to understand trends in the electronic and optical properties that make the molecular dyes suitable for these particular applications. The resources of the Centre for High Performance computing help to facilitate this by making it possible to carry out molecular modelling calculations that can be used to predict how changes to the structures of the molecular dyes will modify their properties. The flexibility that CHPC provides where memory allocations are concerned is often vital.
The INI currently has three staff members, twelve PhD, fourteen MSc and five Honours students, and four postdoctoral fellows, and usually publishes over forty peer-reviewed publications per year. This rose to a high of sixty-five in 2017 with ten involving the use of CHPC resources.
Principal Investigator: Dr Madison Lasich
Institution Name: Mangosuthu University of Technology
Active Member Count: 17
Allocation Start: 2021-02-15
Allocation End: 2021-08-14
Used Hours: 192047
Project Name: Chem Thermo
Project Shortname: CHEM1028
Discipline Name: Chemical Engineering
The Thermodynamics-Materials-Separations Research Group (TMSRG) at Mangosuthu University of Technology employed a multi-scale modelling approach to study the use of waste cement and concrete as a potential medium for a particularly challenging industrial gas separation; namely separating mixtures of ethylene and acetylene. These components are important as feedstocks in many downstream chemical processes, and often occur together, however, they are difficult to separate as they have similar boiling points and the molecules are similar in size. Our study combined simulations at the molecular level with macro-scale modeling of a real industrial separation unit to evaluate the performance of cement hydrate (the hardened form of cement paste) as a medium for this separation at various gas compositions, temperatures, and pressures. Cement hydrate displayed potential usefulness in removing acetylene from ethylene under targeted operating conditions. This work was published in ACS Omega, an international peer-reviewed open access journal [article DOI: 10.1021/acsomega.1c02902].
Principal Investigator: Dr Michael Barnes
Institution Name: South African Weather Service
Active Member Count: 1
Allocation Start: 2021-02-16
Allocation End: 2021-08-15
Used Hours: 259818
Project Name: SAWS Storm Surge and Wave Climatology (ERA5)
Project Shortname: ERTH1395
Discipline Name: Earth Sciences
The SAWS Marine Research Unit have started to complete a 41-year storm surge hindcast for the South African coastline. This will be the first of its kind for the country. Little is known about the storm surge climate of the South African coastline. Applying the numerical modelling system that is used for short-ranged forecasting with long-term climate data forcing allows us to look at past storm surge events over a long period of time. The majority of the hindcast has been completed and some initital analysis and result have been obtained that we hope to publish once completed.
Principal Investigator: Dr Mary-Jane Bopape
Institution Name: South African Weather Service
Active Member Count: 26
Allocation Start: 2021-02-16
Allocation End: 2021-08-15
Used Hours: 644868
Project Name: Very Short Range Forecasting
Project Shortname: ERTH1022
Discipline Name: Earth Sciences
Weather forecasting beyond a lead time of two hours relies on the use of numerical weather prediction models. In Africa these models are often used without much testing and as black boxes. Work done on the CHPC cluster in collaboration with five other countries within SADC allowed us to test the performance of different subgrid models over Southern Africa. This research allowed us to see some shortcomings in available schemes, however the research also made clear challenges associated with limited observations on the continent.
Principal Investigator: Prof Thuto Mosuang
Institution Name: University of Limpopo
Active Member Count: 10
Allocation Start: 2021-02-16
Allocation End: 2021-08-15
Used Hours: 1696
Project Name: Computational studies of various ultra-hard materials
Project Shortname: MATS0875
Discipline Name: Material Science
his is a sensors and energy materials group at the Department of Physics, University of Limpopo. The group has two Doctoral, two Masters students, and two Honours students. Computationally the research projects investigate nanomaterials like copper selenide, gallium nitride, gallium arsenide graphene oxide, and boron nitride. Specifically, electronic, structural, thermodynamic properties are investigated to enhance semi-conductivity. These properties are then mapped with experimental properties for possible gas sensing and energy materials. DL_POLY software through the CHPC is being utilised in the structural, dynamical, and thermodynamic properties. The exciting code also through the CHPC is being utilised in the electronic and excited properties of the materials.
Principal Investigator: Prof Gert Kruger
Institution Name: University of KwaZulu-Natal
Active Member Count: 16
Allocation Start: 2021-02-17
Allocation End: 2021-08-16
Used Hours: 310059
Project Name: CPRU molecular modeling
Project Shortname: HEAL0839
Discipline Name: Health Sciences
Who) The Catalysis and Peptide Research Unit at UKZN (http://cpru.ukzn.ac.za/Homepage.aspx) is currently working on the mechanism of action on the HIV PR with respect to the natural substrates. (What) We use an quantum chemical/molecular mechanics hybrid (Oniom) approach that gives us high level results on the thermodnamics and kinetics of the reaction. Our model enables us to study application of both electronics and sterics to change the bond strenght and binding energies of new potential inhibitors. The same approach is followed for transpeptidases, essential enzymes for the inhibition of TB. (Why) HIV is a major health threat in SA and new drugs is required due to the constant development of drug resistance. Similarly, TB and TB co-infection with HIV is a mojor health issue. (How) Our theoretical studies will allow us to test a computational model that correctly describes the interactions between the respective enzymes and existing drugs. We are improving our computational model to the extent where are now in a position to start proposing new potential drug leads for synthesis. This can only be done with large computational resources provided by the CHPC.
Principal Investigator: Prof Mpho Sithole
Institution Name: Sefaku Makgatho Health Sciences University
Active Member Count: 6
Allocation Start: 2021-02-17
Allocation End: 2021-08-26
Used Hours: 92946
Project Name: Computational modeling of titanium based alloys
Project Shortname: MATS1228
Discipline Name: Physics
The research group is based at Sefako Makgatho Health Sciences University. This is the 3rd year of the group. The group is collaborating with colleagues from CSIR Pretoria. The research projects are based on Titanium alloys [Ti-Mo-Zr] for biomaterial used as bone plates and orthopaedic implants and permanent magnets [Nd-Fe-B] for electronic appliances, generators and electric cars. The project employ the First Principle Method using CASTEP in Material Studio. The calculations are based on DFT. The group rely on CHPC for handling huge data and for high speed data calculations.
Principal Investigator: Dr Sunita Kruger
Institution Name: University of Johannesburg
Active Member Count: 1
Allocation Start: 2021-02-17
Allocation End: 2021-08-16
Used Hours: 516299
Project Name: Environmental Heat Transfer
Project Shortname: MECH0995
Discipline Name: Computational Mechanics
This research group in the Mechanical Engineering Science Department of the University of Johannesburg is currently focusing on environmental heat transfer. Specifically heat transfer in naturally ventilated greenhouses. Greenhouses are used to protect plants from adverse weather conditions and insects. Ventilation of greenhouses are of vital importance to ensure quality crop production. If temperatures in a greenhouse is too high, poor plant growth may result, and an increased need for frequent watering. A mechanical ventilation system might be required to cool the inside of the greenhouse. Natural ventilation is an alternative option used to ventilate greenhouses. Natural ventilation uses temperature and wind to control the indoor climate of greenhouses. Unfortunately greenhouses are extremely energy intensive. Energy costs are the third highest cost related to greenhouse crop cultivation. Reducing the operating costs of energy associated with greenhouse cultivation may result in a price reduction of greenhouse cultivated crops. Conducting experimental work on ventilation of greenhouses can be costly and cumbersome. Using computational methods such as CFD (Computational Fluid Dynamics) to obtain qualitative and quantitative assessment of greenhouses can reduce costs and time involved. The computer cluster at the Centre for High Performance Computing has been used to conduct the numerical investigation using CFD. Currently research is being conducted on a large commercial greenhouse. Smaller greenhouses containing a single span have also been investigated. Currently, three dimensional models of greenhouses subject to buoyancy driven flow are being investigated
Principal Investigator: Dr Jaap Hoffmann
Institution Name: Stellenbosch University
Active Member Count: 5
Allocation Start: 2021-02-17
Allocation End: 2021-08-16
Used Hours: 37327
Project Name: Flow through porous media
Project Shortname: MECH1116
Discipline Name: Computational Mechanics
Prof. Jaap Hoffmann and his students at Stellenbosch University embarked on a CFD - DEM study of heat and mass flow through packed beds in 2017. We use the CHPC to simulate heat and mass transfer in the interstitial volumes. A new group member is looking at the effect of additional turbulence due to the particles on pressure drop and heat dispersion.
Packed bed thermal energy storage is a crucial component of dual cycle solar thermal plants. In order to drive cost down, it is imperative to reduce capital and pumping costs for these beds. CFD simulations enabled us to see detailed flow patterns inside a rock bed, that helped us to use 3D simulations to improve bed design. The flow resistance and heat transfer in a packed bed of irregular particles are highly anisotropic. Conventional (isotropic) models leads to erroneous solutions, and excessive capital or pumping cost.
Two equation modelling of the pressure drop through the bed is complete. In 2021, the focus will be on heat transfer and turbulence generation/dissipation due to particles.
Principal Investigator: Prof Gerard Tromp
Institution Name: Stellenbosch University
Active Member Count: 10
Allocation Start: 2021-02-17
Allocation End: 2021-08-16
Used Hours: 118116
Project Name: South African Tuberculosis Bioinformatics Initiative
Project Shortname: CBBI0999
Discipline Name: Bioinformatics
The South African Tuberculosis Bioinformatics Initiative (SATBBI) performs an important role in tuberculosis research by supporting numerous research projects with data analysis and modeling. SATBBI also has an important educational and training mandate. The CHPC is a vital resource for these activities.
Principal Investigator: Prof Shahida Moosa
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-02-17
Allocation End: 2021-08-16
Used Hours: 62360
Project Name: Rare Disease Genomics
Project Shortname: HEAL1396
Discipline Name: Health Sciences
The Rare Disease Genomics in South Africa research group is based at Stellenbosch University's Faculty of Medicine and Health Sciences. We perform next generation sequencing on samples form participants with undiagnosed rare diseases. Whole exome sequencing has been shown to be an excellent diagnostic tool in international studies. Our preliminary data show that this is certainly the case for South Africa also. With the help of the CHPC, we are able to build and optimize our analysis pipelines and provide diagnoses to the undiagnosed in southern Africa. This is improving our understanding of the genomic basis of disease in our populations. This work will be used to leverage support for broader access to exome sequencing in our setting.
Principal Investigator: Prof Yoshan Moodley
Institution Name: University of Venda
Active Member Count: 6
Allocation Start: 2021-02-17
Allocation End: 2021-08-16
Used Hours: 81316
Project Name: Mammalian Evolutionary Genomics
Project Shortname: CBBI0911
Discipline Name: Bioinformatics
This research group based at the University of Venda in the tropical far north of South Africa, focusses on reconstructing mammalian evolutionary history from whole genome sequences. As far as I am aware, there is less than a handful of other institutions in Africa that are able to carry out this line of research. There are various reasons for this. The last 10 years has seen a massive paradigm shift in biology towards big data research, spurred on by next generation (genome scale) sequencing technologies as well as other omics strategies and the availability of large data crunching servers like the CHPC. RSA, has been slow in both the generation and the analysis of genome wide data, with the greatest advances coming from the medical and agricultural fields, with little development in the zoological and evolutionary contexts. This group breaks that mold. There are currently several projects being run under this theme by my postgraduate and postdoc researchers. Due to the big data nature of these projects, it is absolutely essential that users have access to an external resource like the CHPC, as this not only provides high performance computing capabilities, but also a "safe and reliable" off campus place to run our analyses, which is independent of the dysfunctionality that otherwise characterizes this university.
Principal Investigator: Dr Rian Pierneef
Institution Name: University of Pretoria
Active Member Count: 12
Allocation Start: 2021-02-18
Allocation End: 2021-08-17
Used Hours: 67163
Project Name: Bioinformatic and Computational Biology analyses of organisms
Project Shortname: CBBI1124
Discipline Name: Bioinformatics
The research programme "Bioinformatic and Computational Biology analyses of organisms" is a diverse and dynamic research programme based on the bioinformatic analysis of next-generation sequencing data. This spans various "omics" fields such as genomics, metagenomics, viromics and transriptomics to name a few. Dr Pierneef, PI, is based at the Agricultural Research Council and members of this research programme include other researchers and students. The programme is focused on addressing problems of critical importance in the South African agricultural value such as food borne pathogens, improved breeding, plant disease and soil health. This programme has further progressed to include wastewater-based epidemiology and in particular SARS-CoV-2 as detected in wastewater. The CHPC is a valuable partner in all these research endeavors an serves as the starting point for big data analysis as produced by sequencing technologies. The CHPC provides a critical service as the majority of these projects require large computational resources. This research programme is growing in leaps and bounds and this is largely due to access to the CHPC which ensures data is processed in a proficient and stable manner.
Principal Investigator: Dr Jaco Badenhorst
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 4
Allocation Start: 2021-02-19
Allocation End: 2021-08-18
Used Hours: 2063
Project Name: Automatic transcription of broadcast data
Project Shortname: CSIR0978
Discipline Name: Other
The Voice Computing (VC) Research Group at the CSIR Nextgen Enterprises and Institutions (NGEI) cluster develop speech and language-related technologies for the South African context to enhance access to information and communication. Given the multilingual nature of South Africa, human language technologies (HLTs) can make information and services accessible to a larger proportion of the South African population in a sustainable way, by reaching people in remote locations, people who are not necessarily trained in using technology, or people who are not fluent in English. HLTs can also provide access to information to people with disabilities. HLT can support language diversity and providing of information in multiple languages in an affordable and equitable fashion. The development of all 11 official languages is a national priority, which requires significant attention to HLT in all of these languages. HLT can be of significant economic importance – both by empowering citizens of the country to work more productively, and by forming the core of an exportable set of technologies (since such technologies are currently not available to most of the developing world).
Principal Investigator: Prof Scott Hazelhurst
Institution Name: University of the Witwatersrand
Active Member Count: 4
Allocation Start: 2021-02-22
Allocation End: 2021-08-21
Used Hours: 84134
Project Name: Bioinformatics and Experimental Algorithms (BEAT)
Project Shortname: CBBI0930
Discipline Name: Bioinformatics
The Wits Bioinformatics group based at the Sydney Brenner Institute for Molecular Bioscience at the University of the Witwatersand researches the impact of variations in a group of genes that are very important for drug metabolism, particularly in African populations. We are working with other African researchers as part of the Human Health & Heredity in Africa (H3A) with financial support from GSK.
Variations in these ADME genes may impact on how effective a drug is or mean that certain people may be more likely to have adverse drug reactions when they have these drugs. In our prior work, we discovered a number of novel variations in newly sequenced data from different African countries and we are now trying to understand their impact.
Using the high powered Graphical Processing Units (GPUs) we have been able to study the impact of mutations in three of these genes, CYP3A5, CPY2D6 and G6PD.
Th work is extremely computationally intensive as it requires simulating the behaviour of DNA molecules at very fine resolution as they interact with each o there and drugs molecules.
Some of the variations impact the plasticity of the resulting proteins when interacting with drugs. With further work we hope to bring greater understanding into pharmacogenomics in Africa.
The group has published one paper with further papers under preparation. We expect one PhD and one MSc student to complete by the end of the year.
Principal Investigator: Mrs Chantel Niebuhr
Institution Name: University of Pretoria
Active Member Count: 3
Allocation Start: 2021-02-22
Allocation End: 2021-08-21
Used Hours: 393599
Project Name: Hydrokinetic turbine analysis
Project Shortname: MECH1174
Discipline Name: Other
The Hydro research group at the University of Pretoria Department of Civil Engineering has focussed a large potion of research in developing usable digital twins of hydrokinetic installations in Channel flows through validated CFD models. This allows testing of a broad range of operational variations as well as in depth knowledge on the flow dynamics occurring behind the turbine which is difficult to visualize experimentally. The work is being done with the purpose of developing usable simplified hydrodynamic prediction models of the wake and backwater effects resulting from a turbine installation. This allows more "knowns" around these installations and therefore increases confidence in this new technology. Theses models however require complex 3-dimensional modelling and large meshes which is only possible through the CHPC resources.
Principal Investigator: Prof Andries Engelbrecht
Institution Name: Stellenbosch University
Active Member Count: 16
Allocation Start: 2021-02-22
Allocation End: 2021-09-17
Used Hours: 1231240
Project Name: Computational Intelligence for Optimization
Project Shortname: CSCI0886
Discipline Name: Computer Science
Algorithmic models of social organisms in nature, such as bird flocks, have been developed to solve complex optimization problems. Real-world problems include knowledge discovery from datasets and portfolio optimization. Before these algorithms can be used to solve real-world optimization problems, it is necessary to gain a clear understanding of how they work, and to gauge their performance on benchmark problems, in comparison with established algorithms. The empirical analyses required is computationally expensive, and take excessively long on standard desk top computers. The CHPC provides a means to reduce the time to quantify the performance of these algorithms prior to deployment to solve real-world problems.
Principal Investigator: Dr Annerine Roos
Institution Name: University of Cape Town
Active Member Count: 7
Allocation Start: 2021-02-23
Allocation End: 2021-08-26
Used Hours: 1963
Project Name: DCHS and MRI studies
Project Shortname: HEAL1267
Discipline Name: Health Sciences
The aim of the Drakenstein Child Health Study is to investigate determinants of early child development in a cohort of >1000 mother-child dyads of the Drakenstein district in the Western Cape of South Africa. The communities participating in the study represent a low-middle income country setting that is characterised by high substance abuse, poverty and violence. The Brain imaging subgroup include international collaborators of the Universities of Cape Town, Stellenbosch and Los Angeles, California. We aim to assess longitudinally cognitive, behavioural and neural outcomes in children from birth to 6 years of age, to identify vulnerable groups and key periods for intervention that may optimise development. Evidence show the earlier the interventions, the better the outcomes.
CHPC resources are used by this programme to analyse brain imaging data including structural data, of children with prenatal exposure to maternal depression and substances (e.g. alcohol and tobacco smoking). Prenatal exposure to maternal depression and/or substances may adversely affect the underlying neural pathways, that may present as impaired cognitive and behavioural development in offspring.
Principal Investigator: Prof Johan W. Joubert
Institution Name: University of Pretoria
Active Member Count: 3
Allocation Start: 2021-02-23
Allocation End: 2021-08-22
Used Hours: 98913
Project Name: Mobility modelling using vehicle telematics
Project Shortname: MECH0803
Discipline Name: Other
Government (all spheres) is tasked with the challenge of deciding where to invest in infrastructure (for housing/transport/economic development etc). The real hard part is to balance two questions. Firstly, "who gets the benefit of infrastructure?" and, secondly, "who PAYS for those benefits?". In our research programme, we develop agent-based transport models that are able to depict and imitate the complex, multi-modal and large-scale transport scenarios in a South African context. Each individual - from the Metrorail commuter to the private car owner, to the minibus taxi driver, to the commercial vehicle delivering your goodies at the Woolies Foods - are represented in such a transport model. The goal of our ongoing research is to assist government and industry make better informed, evidence-based decisions about the intended, and unintended consequences of specific transport interventions. Since every individual is modelled the computational burden is substantial and the CHPC allows us to overcome the data preparation and simulation challenges of executing the large-scale, MATSim models.
Principal Investigator: Ms Ione Loots
Institution Name: University of Pretoria
Active Member Count: 1
Allocation Start: 2021-02-24
Allocation End: 2021-08-23
Used Hours: 502059
Project Name: Flow patterns in service reservoirs
Project Shortname: MECH1401
Discipline Name: Other
Reservoir design has a major impact on the quality of the water being provided to the consumer. The maintenance of acceptable water quality in service reservoirs relies heavily on effective mixing. Consequently, the presence of stagnant water zones within these reservoirs gives rise to significant alarm. These stagnant zones, commonly known as dead zones, are characterized by sluggish or non-existent water flow, presenting a significant issue. In the study, flow patterns in a typical circular service reservoir are investigated using Computational Fluid Dynamics (CFD). CFD is a very specialised field of numerical methods to simulate and analyse fluid flow. CFD simulations can provide valuable insights into fluid flow behavior, such as velocity distributions, pressure gradients, and turbulence characteristics. The computational requirements of CFD stem from the inherent complexity and size of the governing equations. Simulating real-world scenarios often involves modeling complex geometries and intricate boundary conditions. These simulations require significant computational resources, such as high-performance computing clusters to handle the number of calculations. Therefore, this study heavily relies in the CHPC for its high performing computing to analyse the reservoir flow.
The research is carried out with a combined effort from the Civil Engineering Department and Microbiology Department at the University of Pretoria. The Civil Engineering Department carries out the computing aspects of the study, giving guidance to the Microbiology Department of where to sample water to analyse the quality of the water in the reservoir.
Principal Investigator: Prof Yin-Zhe Ma
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-02-24
Allocation End: 2021-09-16
Used Hours: 10023
Project Name: Machine learning in 21-cm cosmology
Project Shortname: ASTR1323
Discipline Name: Astrophysics
The purpose of using the CHPC is to manipulate the hydrodynamic simulation "IllustrisTNG" to understand the dispersion relation of Fast Radio Bursts from Intergalactic medium. We utilise the
datasets from this simulation and FRB observation to present an alternative method that could be used to extract photo-ionization efficiency (how hydrogen is being ionised at low redshifts) in a wide redshift range. Results obtained from this method could be used to cross-check those obtained from Lyman-alpha forest observations. CHPC is critical for this study because manipulating the large Illustris simulation requires a lot of memories and threads in parallel.
Principal Investigator: Dr Frederick Malan
Institution Name: University of Pretoria
Active Member Count: 1
Allocation Start: 2021-02-24
Allocation End: 2021-09-02
Used Hours: 49642
Project Name: Proton-Responsive Metal Complexes as Catalysts for Sustainable Hydrogen Production and Storage
Project Shortname: CHEM1358
Discipline Name: Chemistry
The research of the Malan Inorganic Chemistry Research group, led by Dr Frederick Malan (at the University of Pretoria), revolves around the design, economic synthesis, and characterization of new transition metal compounds. The (homogeneous) catalytic activity of these compounds are then evaluated in a variety of oxidation and reduction reactions. Additional electrochemistry, computational chemistry and single crystal X-ray diffraction studies aid in the understanding of the working of these catalysts. The investigation of the role of proton-responsive ligands in CO2 and N2 fixation processes, to efficiently make use of cheap building blocks as an entry point to fuel and other fine chemicals, and to curb the effect of greenhouse gas pollution, remains important. Catalysts reduce the activation energy by which a reaction takes place, and therefore the calculation and prediction thereof (using the CHPC) is important in this research. To date, promising results have been obtained with three honours projects, one Masters project, as well as more results (especially from a computational point of view) anticipated for the remainder of 2021 as one new honours and one new PhD student initiated his research project.
Principal Investigator: Prof Zander Myburg
Institution Name: University of Pretoria
Active Member Count: 6
Allocation Start: 2021-02-25
Allocation End: 2021-09-02
Used Hours: 112378
Project Name: Forest Molecular Genetics (FMG) Programme
Project Shortname: CBBI1015
Discipline Name: Bioinformatics
The Forest Molecular Genetics (FMG) Programme at the University of Pretoria focuses on the genetic control of growth and development in fast-growing plantation trees with a view to enhance biomass production and improve wood properties for timber, pulp, paper, and biomaterials production. They work in close collaboration with South African forestry companies through the FMG Consortium (FMGC) to develop capacity and resources for the application of tree biotechnology in operational tree improvement programmes. The group has successfully used a systems genetics approach in an interspecific backcross population of E. grandis x E. urophylla to map key genomic regions affecting gene expression and metabolic profile associated with variation in growth and wood chemistry. They recently embarked on an effort to engineer cell wall traits such as xylan content and structure affecting pulp yield employing new approaches such as CRISPR-Cas9 genome editing. Over the past five years FMG have successfully used a single nucleotide polymorphism (SNP) marker chip with 60,000 DNA markers to genotype over 3000 Eucalyptus trees and recently also established a multi-species SNP genotyping chip for tropical pines. These SNP chips will aid in genome-assisted breeding projects in Eucalyptus and pine where the aim is to fast-track breeding cycles. The programme is also constructing a Genome Diversity Atlas for Eucalyptus and pine species grown in South Africa, laying the foundation for the emerging field of landscape genomics, which combines population genomics with analysis of interactions with environmental factors. Towards this, FMG is undertaking an effort to use long-range single-molecule sequencing to sequence genomes of several hundred Eucalyptus trees, also expanding to other genera including acacia, macadamia and pine. The programme has a strong collaborations with the Bioinformatics and Computational Biology Centre at the University of Pretoria where we access to a few large servers and a cluster. However, there is a need for additional resources, specifically for projects that rely on high-throughput processing of sequence data.
Principal Investigator: Dr Brigitte Glanzmann
Institution Name: Stellenbosch University
Active Member Count: 18
Allocation Start: 2021-02-25
Allocation End: 2021-08-24
Used Hours: 297550
Project Name: SAMRC Precision Medicine African Genomics Centre
Project Shortname: CBBI1195
Discipline Name: Bioinformatics
Next generation sequencing (NGS) technologies on the African continent have become even more important throughout the Covid-19 pandemic, where shipping of samples to overseas laboratories and service providers has become almost impossible. NGS technologies however lead to the generation of large data sets which require extensive bioinformatics and computational resources. Computational life sciences therefore relies on the implementation of well-structured data analysis pipelines as well as high-performance computing (HPC) for large-scale applications. Here, we report the sequencing of the numerous samples at the South African Medical Research Council's Genomics Centre, and the processing of the data in collaboration with the Centre for High Performance Computing (CHPC). This has proven particularly important in terms of the Covid-19 pandemic, where genetic investigations into severe Covid-19 as well as Covid-19 infections in children. The active collaboration with the CHPC has ensured that this groundbreaking research can continue.
Principal Investigator: Prof Joanna Dames
Institution Name: Rhodes University
Active Member Count: 7
Allocation Start: 2021-02-25
Allocation End: 2021-09-02
Used Hours: 36794
Project Name: Mycorrhizal Interactions
Project Shortname: CBBI0966
Discipline Name: Bioinformatics
The Mycorrhizal Research group is based at Rhodes University, Grahamstown.
Mycorrhizal fungi form a symbiotic relationship with the roots of the majority of plant species. The interaction between these soil fungi, other soil microbes and plants results in many beneficial growth effects making an important contribution to sustainable agriculture, horticulture, and environmental rehabilitation. There are several types of mycorrhizal relationships depending on the fungi and host plants involved. Little is known about the biodiversity of these fungi in South African soils. The use of next-generation sequencing provides an opportunity to unravel this biodiversity in order to better understand factors that impact the relationship. The CHPC platform has assisted in the analysis of biodiversity of fungi in general and mycorrhizal fungi in particular as well as some interactions with soil bacteria.
Principal Investigator: Prof Francois Engelbrecht
Institution Name: University of the Witwatersrand
Active Member Count: 3
Allocation Start: 2021-02-26
Allocation End: 2021-08-25
Used Hours: 408886
Project Name: WITS Climate Modelling
Project Shortname: ERTH1200
Discipline Name: Earth Sciences
The Intergovernmental Panel on Climate Change (IPCC) Assessment Report Six (AR6) Working Group I report was published on 9 August 2021. Several chapters of the report are devoted to regional climate change. The report makes reference to no less than seven regional climate modelling papers using the CCAM model (simulations performed on the CHPC clusters, under my leadership, across various allocations and simulation cycles). The CHPC, most recently through its Lengau cluster, has thus directly contributed to the most authoritative climate report in the world.
Within the current cycle, testing of the new CCAM code on the Lengau cluster has been completed. The programme at the Wits GCI now moves to a more computationally intensive phase, namely downscaling of the CMIP6 GCMs.
Principal Investigator: Dr Samson Khene
Institution Name: Rhodes University
Active Member Count: 3
Allocation Start: 2021-02-27
Allocation End: 2021-08-26
Used Hours: 7765
Project Name: Synthesis, Spectroscopy and Nonlinear Optical Properties of Phthalocyanines
Project Shortname: CHEM0975
Discipline Name: Chemistry
Many of today's optical sensors are based on optical materials in which non-linear optics may play a role. Defects in the cable can have an effect of reducing the light (which carries information) in an optical fiber or cable. This phenomenon is also present in a variety of non-linear active materials or molecules. This nonlinear reduction of light phenomenon can be used as a sensing method for detection and correction of signals that are sent in cables. The aim of our research (at Rhodes University in the Chemistry department) is to make and study molecules which will help improve signal transmission using a nonlinear process. This work will not be possible if it was not supported by CHPC facilities.
Principal Investigator: Prof Krishna Bisetty
Institution Name: Durban University of Technology
Active Member Count: 3
Allocation Start: 2021-02-27
Allocation End: 2021-08-26
Used Hours: 3952
Project Name: Computational Modelling & Bioanalytical Chemistry
Project Shortname: CHEM0820
Discipline Name: Chemistry
Our current in silico work used to supplement the electrochemical experiments has now been recognized by leading Food Science Journal as a new way to investigate the interaction between aptamer and OTA. The impact of this work, especially the molecular modelling and simulation studies would directly contribute to the development of human capacity development in SA. Young Black postgraduate students will have hands-on experience with the state-of-the-art simulation protocols.
Principal Investigator: Dr Chris Spies
Institution Name: Agricultural Research Council
Active Member Count: 1
Allocation Start: 2021-03-01
Allocation End: 2021-08-28
Used Hours: 6192
Project Name: Fungal diversity and taxonomy
Project Shortname: CBBI1397
Discipline Name: Bioinformatics
The Soilborne Plant Diseases Unit of the Agricultural Research Council – Plant Health and Protection (ARC-PHP) investigates soilborne pathogens of crops and the management of diseases caused by them. This contributes to food safety and security by improving crop yields and quality. Current projects include diseases of a important crops such as maize, soybean, potatoes, and olives.
Research projects at this unit typically include the identification of fungi from soil and plant roots, trials to determine the impact of individual fungi on plants, and trials to identify treatments that limit the negative impact of pathogens. In this context, species-level identification of fungi are important for several reasons. Firstly, the soil and root environments harbour diverse populations of fungi, of which many are beneficial to plants. Individual fungal genera may actually include pathogenic and beneficial species. Secondly, different pathogens can cause similar diseases. This implies that the organism causing the disease cannot be identified based on observed symptoms. Thirdly, treatments may vary in their efficacy toward different pathogens. Consequently accurate identifications are required to reliably diagnose problems, and recommend effective solutions to crop industries and producers.
Accurate species identification of fungi require DNA sequencing of specific gene regions that can discriminate between different species (i.e. barcoding regions). DNA sequences are aligned with reference sequences and evolutionary relationships of the included taxa are inferred using phylogenetic analyses to identify species. In some cases more than one gene region is needed to delineate different species. In genera with many species, this can lead to large, complex datasets that cannot be analysed on standard desktop or laptop computers, and for which the CHPC is used to conduct phylogenetic analyses.
Once reliable species identities have been obtained, representative fungal strains are deposited in the National Collection of Fungi for safekeeping, thereby contributing to our National Assets.
Principal Investigator: Prof Phuti Ngoepe
Institution Name: University of Limpopo
Active Member Count: 27
Allocation Start: 2021-03-02
Allocation End: 2021-09-22
Used Hours: 4726209
Project Name: Computational Modelling of Materials: Energy Storage
Project Shortname: MATS0856
Discipline Name: Material Science
The Research Group is based in Materials Modelling Centre at the University of Limpopo, under the leadership of Professor Phuti Ngoepe. Energy Storage Systems: We have, through simulations on primary particles of LMO spinel, characterized disruptive transition during discharging; and are currently exploring porosities to enhance capacity of such cathodes. All manganese NMC primary nanoparticles have been modelled and modified to NMC and the nature of structural disorder associated with discharging are studied. This will inform the new manganese rich NMC co-precipitation experiments set up on the UL campus. In addition, stable O3 structures with enhanced performance, are predicted by doping using 4IR machine learning methods. Approaches of minimizing anion oxidation at surfaces during discharging are explored. Some aspects of the beyond lithium ion batteries are continuing in the form of Li-S,Se, particularly on the S rich side, together with work on catalysts for Li,Na,K,-air batteries. Mineral processing: After the development on proof of concept pertaining to agreement of simulations and experiments in the design on reagents for mineral recovery from complex ore, we have proceeded with the applications to precious metal chacogenides, where good experimental samples are not readily available. We are exploring applicable reagents. We have also studied new reagents that are applicable for copper extraction and are identifying promising candidates. Alloy development: We have continued with phase stability studies of shape memory alloys using a of combination of energetics, elastic and vibrational properties. The extension of the study to TiPd alloys is now advanced. Semi-empirical methods are being employed to explore phase transformations at elevated temperatures. Global minimisation methods study for nanoclusters of Ti atoms are continuing.
The computational modelling studies, through CHPC, are intended to predict processes that are essential in pilot and production plants for advanced battery systems, mineral processing and metal production industries.
Thus far the progress on such studies is excellent and whilst delivering products it is developing essential human capacity.
Principal Investigator: Dr Ouma Moro
Institution Name: North-West University
Active Member Count: 5
Allocation Start: 2021-03-02
Allocation End: 2021-09-09
Used Hours: 267593
Project Name: Rational design of novel catalysts using Chemistry 4.0 and 4IR tools
Project Shortname: MATS1365
Discipline Name: Material Science
Chemistry 4.0 is as at the core of material's informatics in the Forth Industrial Revolution (4IR) space. Borrowing from the affordances of the materials genome initiative, machine learning (ML) with(out) deep learning (DL) techniques with neural networks can be used to design the next-generation catalyst(s) for hydrogen evolution reactions, oxygen evolution reactions and hydrogenation and dehydrogenation reactions. At the core of this project is the application of density functional theory calculations to calculate descriptors that will used as inputs to machine learning models and also for training the ML and DL models. This will facilitate the rapid scaling of material design and design within the continent and also the Republic of South Africa. This because South Africa is well endowed with mineral resources essential for energy storage and catalysis and as governed by the Energy White Paper, this project offers innovative applications and ideas on how the government can scale up technologies related to energy as well as build capacity with the country and continent. Those working in the project include: Dr Cecil NM Ouma (the Principal Investigator with vast experience in ab initio modelling), Gladys Kingóri (A PhD student at Technical university of Kenya) and Hezbon Ondieki (A prospective PhD student). The HPC resources from CHPC in Cape Town has been at the core of all the research that we have done so far. It is these resources that are both facilitating and mediating Africa's concerted efforts towards catching up with developed nations in terms of research, development and capacity building.
Principal Investigator: Dr Abdulrafiu Raji
Institution Name: University of South Africa
Active Member Count: 0
Allocation Start: 2021-03-03
Allocation End: 2021-09-09
Used Hours: 497469
Project Name: Magnetocaloric effect in selected metallic nanoclusters on two-dimensional (2D) substrates, and in selected rare-earths.
Project Shortname: MATS1156
Discipline Name: Material Science
The research study is based at the Department of Physics of the University of South Africa (UNISA) and is being led by Dr. Abdulrafiu Raji. The study concerns computational studies of two-dimensional (2D) nanomaterials for potential application in spintronics, ultrathin refrigeration and air-conditioning systems and any nanosystems where cooling may be needed. Specifically, the research aims to study electronic and magnetic interactions as well as magnetocaloric effect (MCE) in metallic nanostructures deposited on various two-dimensional substrates, such as graphene, silicene, and similar 2D materials. Magnetocaloric effect is a property of magnetic materials which could be exploited for refrigeration purposes. While the conventional refrigerator system is based on compression and evaporation of often environmentally unfriendly gases, magnetic refrigeration systems is based on magnetizing and demagnetizing a magnetic material. In the last few years, the focus in magnetocaloric research seems to have shifted toward ultrathin materials. Because of plethora of candidate materials that may be investigated as possible candidates for MCE, performing experimental investigations may be time and resources consuming. This computational studies could provide the lead and narrow the range of materials that could be considered for further experimental investigations. Performing ab-initio density-functional theory (DFT) study to determine the properties of candidate material systems is a viable and state-of-the-art approach which can complement experimental efforts. DFT calculations require specialized softwares as well as high-capacity data storage and memory requirements which are often beyond the capability of desktop computers. Thus, the availability of high-performance computing (HPC) facility is absolutely necessary.
Principal Investigator: Dr Adeniyi ogunlaja
Institution Name: Nelson Mandela Metropolitan University
Active Member Count: 5
Allocation Start: 2021-03-04
Allocation End: 2021-08-31
Used Hours: 1847
Project Name: Chemistry and Material Science
Project Shortname: MATS1070
Discipline Name: Chemistry
Dr Adeniyi Ogunlaja (Research group -Analytical/Inorganic) at Nelson Mandela University. Refractory polyaromatic hydrocarbons found in fuels such as dibenzothiophene, quinoline and its alkylated derivatives emits nitrous oxides and sulfur oxides to the environment when combusted, thereby reducing air quality. Desulfurization by means of oxidative desulfurization and adsorptive denitrogenation has been proposed to complement the hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) processes, currently employed to remove sulfur and nitrogen compounds in fuels. At Nelson Mandela University analytical chemistry research group, we design materials that can selectively remove these sulfur and nitrogen compounds. Molecular modelling based upon the density functional theory (DFT) was employed to understand the mode of adsorption (interaction) between synthesised materials and the various compounds to be removed from fuels. Computational study assists with the understanding of the electronic and thermodynamic properties of synthesised materials prior to application. It also gives an insight as to the specific properties to be targeted when designing adsorbents for environmental clean-up. So far, progress has been made around the design and adsorption of some toxic molecules in fuels. HPC assist our group to build models that could be employed in explaining complex phenomena. It has also enabled us to ask big questions around activity of our materials and to test for possible answers computationally. CHPC has created the platform for where researchers like me could assess the HPC and also be assisted when encountering computation problems.
Principal Investigator: Prof Michelle Kuttel
Institution Name: University of Cape Town
Active Member Count: 4
Allocation Start: 2021-03-05
Allocation End: 2021-09-01
Used Hours: 290117
Project Name: Molecular modelling of microbial polysaccharides and glycoconjugates
Project Shortname: CHEM1242
Discipline Name: Chemistry
We focus on molecular modelling to inform the development of more effective drugs and vaccines against infectious disease. This work is entirely computational and cannot be performed without access to supercomputers.
Principal Investigator: Dr Rose Modiba
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 5
Allocation Start: 2021-03-05
Allocation End: 2021-09-01
Used Hours: 72930
Project Name: Computational modelling of light metals
Project Shortname: MATS1097
Discipline Name: Material Science
The Advanced Materials Engineering group under the Manufacturing cluster at CSIR who are working on the stability of the materials when alloyed with other transition metals would like to thank CHPC for allowing us to use their facilities. The systematic search for novel materials is hardly achievable only via experimental discovery due to the very large combinatorial space spanned by the crystal structures and chemical compositions for compounds. This difficulty is also connected with eventually high and unstable costs of raw materials on the world market and the time-consuming procedures of synthesis and characterization. It is of great importance for the theoretical search for materials to gain detailed insight into the most relevant physical mechanisms which determine the intrinsic properties of the materials. The CHCP is a very good platform for anyone who would like to do simulation on their materials. Thank you CHPC for the workshops provided to equip us with the most relevant and recent technologies to be able to understand our materials.
Principal Investigator: Prof Obodo Kingsley
Institution Name: North-West University
Active Member Count: 6
Allocation Start: 2021-03-05
Allocation End: 2021-09-09
Used Hours: 1316222
Project Name: Ab initio modelling of liquid organic hydrogen carriers catalyts and 2D materials
Project Shortname: MATS1366
Discipline Name: Physics
The current research programme "MATS1366: Ab initio modelling of liquid organic hydrogen carriers catalyst and 2D materials" was set up to carry out research activities in the area of catalyst development and other two-dimensional materials. The principal investigator is based in HySA infrastructure center of competence in the North West University, South Africa. To date, the members of the research program include: Kingsley Obodo, Hailouf Houssam, Ijeoma Onyia, and Stella Ogochukwu. The use of the CHPC facility is enabling the members to undertake in their Masters as well as Ph.D research work. The group members are undergoing significant progress in their work as demonstrated by the novel research generated, which is currently under-review and submitted for publication. Some of the work published during the reporting period are Sn3C2 monolayer with Transition Metal Adatom as well as the dopant complexes and the pristine Ti3N2 MXene monolayer using spin-polarized density functional theory calculations with van der Waals correction. Other work under review and to be submitted include but not limited to the catalyst for liquid organic hydrogen carriers (LOHCs) de-hydrogenation as well as catalysts for water electrolysis. Water electrolysis entails the breaking down of water to give hydrogen and oxygen. The main technique applied is density functional theory as implemented in CASTEP and Quantum Espresso packages.
Principal Investigator: Dr Thishana Singh
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-03-05
Allocation End: 2021-09-01
Used Hours: 37622
Project Name: Combined kinetics,quantum-chemical investigation of reaction mechanisms involving catalysts
Project Shortname: CHEM0821
Discipline Name: Chemistry
The TS Computational Chemistry Research Group (University of Kwazulu-Natal, School of Chemistry and Physics) currently, uses computational chemistry as a tool in the field of Green Chemistry to compare the chemical reactivity properties of bioactive molecules for drug delivery. Conceptual DFT which is often referred to as 'computational nutraceutics' is a concept used to predict the molecular structure, spectroscopy, and chemical reactivity of nutraceuticals (food or part of a food that provides medical/health benefits, including the prevention/treatment of diseases) by means of computational chemistry and molecular modelling. Using this methodology, the aim is to perform a combined electronic and structure investigation to understand enzyme-inhibitor interactions that can aid in the synthesis and strategic design of novel drugs for the treatment of various infectious diseases. The CHPC provides the facilities: hardware, in the form of the Lengau cluster and software programs such as Gaussian16, amongst many others.
Principal Investigator: Dr Saheed Sabiu
Institution Name: Durban University of Technology
Active Member Count: 9
Allocation Start: 2021-03-08
Allocation End: 2021-09-04
Used Hours: 84640
Project Name: Drug Discovery & Development and Viral Metagenomics
Project Shortname: HEAL1361
Discipline Name: Bioinformatics
The group is researching into the detailed concepts of therapeutic mechanisms of secondary metabolites in degenerative and microbial diseases while reporting health benefits in a way that will provide valuable data which will lead to new small molecule targets that could potentially be lead for drug discovery. To achieve this goal, access to CHPC is crucial for the molecular dynamics aspects of our projects.
Besides drug discovery interests, the group is also focusing on molecular dynamics of enteric and respiratory viruses using whole genome sequencing and metagenomic approaches.
Principal Investigator: Prof Bettine van Vuuren
Institution Name: University of Johannesburg
Active Member Count: 5
Allocation Start: 2021-03-08
Allocation End: 2021-09-04
Used Hours: 2353
Project Name: Biocomplexity on sub-Antarctic islands
Project Shortname: CBBI1153
Discipline Name: Bioinformatics
- We are a research group at the Centre of Ecological Genomics and Wildlife Conservation based at the University of Johannesburg, South Africa.
- We are conservation geneticists and are interested in studying spatial and temporal genetic patterns in multiple organisms (plants and invertebrates) on sub-Antarctic islands, with a special focus on Marion Island. We aim to investigate genetic patterns and structure in the context of environmental changes (for example, climate change). We intend on using our study species as proxies for monitoring global change and its impact on biodiversity as a whole.
- We are well aware of the concept of environmental change (such as climate change, which is more pronounced in the sub-Antarctic region). As conservation geneticists, we will investigate the impact that change has on biodiversity. Understanding genetic patterns is crucial, and our results will bring about far-reaching implications for the development of conservation management programs for Marion Island and on a global scale too.
- To do this, we have generated next-generation sequencing (NGS) data using various sequencing chemistries. We will use the CHPC cluster to analyse our NGS data to answer transcriptomic questions, and phylogenomic and population genomic related analyses. The above mentioned data is exceptionally large, and therefore we will not be able to conduct our research without the CHPC resources since no other platforms can handle these large datasets.
- We are in the process of completing three research articles to be submitted for publication soon. In the near future, and based on the work from Daniela and Shilpa, we hope to publish several articles next year.
We are happy with the progress we have made thus far and thank the CHPC for this (the CHPC has been acknowledged in the publications).
Principal Investigator: Prof Matthew Adeleke
Institution Name: University of KwaZulu-Natal
Active Member Count: 12
Allocation Start: 2021-03-09
Allocation End: 2021-09-05
Used Hours: 95815
Project Name: Computational and Evolutionary Genomics
Project Shortname: CBBI1231
Discipline Name: Bioinformatics
The Quantitative and Computational Genomics in the School of Life Science, University of KwaZulu-Natal is using genomics approach to understand how probiotics modulate the rumen microbiome for improved health and productivity of small stock in South Africa. Furthermore, our group seeks to understand genetic and antigenic diversity of Apicomplexan parasites of veterinary and medical importance for molecular diagnosis and designing anticcodial vaccines and in general control measures for improved animal welfare and increased productivity. Part of the study's focus is to utilize computational approach to understand selection signature and explore comparative genomics to vaccine candidates. Samples are usually collected for DNA extraction. Genetic markers of interest are then amplified and the product is then sent for sequencing. The CHPC platform is then utilized for analysing Next Generation Sequencing data to provide interpretation.
Principal Investigator: Dr Gavin Gouws
Institution Name: SAIAB
Active Member Count: 10
Allocation Start: 2021-03-09
Allocation End: 2021-09-05
Used Hours: 30761
Project Name: RADseq studies of aquatic biodiversity
Project Shortname: CBBI1274
Discipline Name: Other
The Marine Fish Research Team in the Aquatic Genomics Research Platform at the National Research Foundation - South African Institute for Aquatic Biodiversity employs a variety of genetic tools to understand the evolution and ecology of marine fishes (and marine invertebrates). We are interested in the relationships between species, populations and individuals, and in the movement and migration of individuals, with a view to informing conservation and management. Over the last year, we have conducted studies examining whether the genetic diversity of commercially-exploited marine fishes have changed over time, as a result of harvesting, and whether an endangered line-fish exists as and should be managed as a single population along the South African coast. For the former, SNP genotype data were generated and analysed using applications on the CHPC, while the latter used the CHPC and it's applications to model patterns of migration, dispersal and changes in population size using mitochondrial and nuclear genetic data. Some of this research has recently been published, with additional manuscripts currently being prepared.
Principal Investigator: Prof Marlo Moller
Institution Name: Stellenbosch University
Active Member Count: 5
Allocation Start: 2021-03-09
Allocation End: 2021-09-05
Used Hours: 38774
Project Name: Tuberculosis Host Genetics
Project Shortname: HEAL1360
Discipline Name: Health Sciences
Although approximately 25% of the world's population is infected with Mycobacterium tuberculosis, the vast majority will never develop any clinical tuberculosis (TB). Inter-individual variation in the immune response plays a major role in determining different clinical outcomes in infected persons. The Tuberculosis Host Genetics group based at Stellenbosch University aim to identify these unknown host genetic factors influencing immunity. Our investigations of the genetic contribution of the human host to individual and population susceptibility to tuberculosis span the full continuum of TB susceptibility, ranging from those extremely susceptible to disease to those seemingly resistant to infection. We have focused on TB genome-wide associations studies, host genes and strain interactions, the role of sex and ancestry in disease, TB "resisters", tuberculous meningitis and primary immunodeficiencies (PIDD) presenting with an increased TB susceptibility. Our experience in the genetic diagnosis of PIDD has allowed us to expand to the diagnosis of other rare genetic diseases by using the pipelines initially developed for PIDD. We also take the work to the cellular level by studying the function of the TB susceptibility genes discovered, with a specific interest in autophagy. Our insights into the latest genetic analyses available have allowed the work on genetic susceptibility to TB to benefit from population genetic analyses and next generation sequencing. Our work leverages the complex ancestry of South Africans, especially the unique genes from KhoeSan ancestry, to find novel genes and pathways involved in TB resistance or susceptibility, as shown by our admixture mapping. We are currently supervising a ground-breaking project on the sequencing and de novo assembly of an admixed South African genome and establishing recombination maps to African groups. Our chief strength is on putting a uniquely African emphasis on the current highly computational field of genetics as it impacts on human disease, while exploring the downstream functional effects. The findings of our work could in future contribute to the development of host-directed therapies. Moreover, we are contributing to diagnosing rare diseases, including primary immunodeficiencies, in a developing country.
Principal Investigator: Prof Adrienne Edkins
Institution Name: Rhodes University
Active Member Count: 1
Allocation Start: 2021-03-09
Allocation End: 2021-09-05
Used Hours: 93975
Project Name: Molecular and Cellular Biology of the Eukaryotic Stress Response
Project Shortname: CBBI1405
Discipline Name: Health Sciences
The Biomedical Biotechnology Research Unit (BioBRU) is a biomedical research unit at the Department of Biochemistry and Microbiology at Rhodes University. We are a collaborative research unit focused on understanding the structure and function of the cellular stress response. The current study identifies new drug targets in Mycobacterium tuberculosis, the causative agent of tuberculosis (TB) disease. We are using in silico modelling to study the native and mutant versions of clinically relevant TB proteins to examine if such mutations destabilize the protein. We hypothesize that our proposed new drug targets are required for stabliising mutant proteins associated with TB drug resistance. Considering that this analysis is computationally resource-demanding, the CHPC platform is invaluable to our study.
Principal Investigator: Mr Anban Pillay
Institution Name: University of KwaZulu-Natal
Active Member Count: 7
Allocation Start: 2021-03-11
Allocation End: 2021-09-16
Used Hours: 144724
Project Name: Deep Learning Approaches
Project Shortname: CSCI1078
Discipline Name: Computer Science
The Deep Learning research group belongs to the Centre for Artificial Intelligence Research Node at the University of KwaZulu-Natal. The Centre for Artificial Intelligence Research (CAIR) (CAIR.za.net) is a South African distributed Research Network that conducts foundational, directed and applied research into various aspects of Artificial Intelligence. The research group conducts foundational research in machine learning in general and deep learning in particular. Deep learning projects require substantial computing resources and the CHPC is vital to its success. The projects involve various area of deep learning that include deep reinforcement learning, convolutional and recurrent neural networks. Application areas include image processing, natural language processing and speech recognition. Several students are busy with projects at present.
Principal Investigator: Prof Phuti Ngoepe
Institution Name: University of Limpopo
Active Member Count: 9
Allocation Start: 2021-03-12
Allocation End: 2021-09-08
Used Hours: 3383663
Project Name: Computational Modelling of Materials: Mineral Processing
Project Shortname: MATS1404
Discipline Name: Material Science
The minerals processing group at the University of Limpopo focuses on using computational modelling methods to study the reactivity of mineral surfaces with reagents for the separation of minerals. This also extends to the derivation of potentials for such minerals in order to use the molecular dynamics methods, where temperature and pressure effects are investigated. The computing of the work requires supercomputers to generate results. This is because the surfaces and molecule dynamics are heavy on small computers and as such a large cluster such as CHPC is of importance to have this research work done. These research works are compared with experiments to find a correlation with the computational work. The research work has so much impact in improving the recovery of the platinum group minerals in South Africa. Currently, there is so much work that has been generated and good correlations has been established between the computational work and experiments, which are supported by publications.
Principal Investigator: Dr ADEDAPO ADEYINKA
Institution Name: University of Johannesburg
Active Member Count: 6
Allocation Start: 2021-03-12
Allocation End: 2021-09-08
Used Hours: 265008
Project Name: Computational Design of Molecules
Project Shortname: CHEM1221
Discipline Name: Chemistry
The Computational Design of Molecules research group is based at the University of Johannesburg. We aim to design new and improved molecules and materials for applications in Catalysis and renewable
energy. Since the availability of energy is one of the main challenges of the
African continent, being able to achieve our aims as a group will provide clean energy solutions for the continent. We use computational chemistry software to
explore the properties of molecules which is responsible for their activity and then use the knowledge gained to design more efficient and improved materials
Principal Investigator: Dr Pranisha Soma
Institution Name: Agricultural Research Council
Active Member Count: 2
Allocation Start: 2021-03-16
Allocation End: 2021-09-12
Used Hours: 4924
Project Name: Cross Breeding
Project Shortname: CBBI1403
Discipline Name: Other
A crossbreeding program utilizing South African cattle breeds was initiated by Prof Scholtz from the Animal Breeding and Genetics department at the Agricultural Research Council – Animal production campus. This project aims to investigate to study the impact of climate change on the various cattle crossbred genotypes. The project is focussed on studying the crossbred animal performance in the presence of climate change, such as the impact of heat stress on the animal's growth. Based on observations from Pyoos (2018), it was noted that the Nguni and Bonsmara crosses (Nguni x Bonsmara and Bonsmara x Nguni) displayed different levels of performance when kept in the same environment. This led to questioning the genetic changes that could play a role in the Nguni and Bonsmara crosses allowing for adaptation. Therefore, it was imperative to analyse the crossbred genomes in order to identify genetic variations that are present in the two crosses. With the use of whole genome sequencing, a clear understanding of the two crosses will be possible as it will provide information on the genes that are present within each cross as well as its heritance patterns from the parent breeds. Large amounts of sequence data is generated, so the use of the CHPC is vital for the processing of the data. Without the CHPC, the data could not be processed using the normal desktop computing system, as it does not have the space nor the computing power to analyse the data. Currently, substantial progress has been made regarding the processing of the raw whole genome sequence data into workable file formats that will undergo further bioinformatics analysis.
Principal Investigator: Dr Henry Otunga
Institution Name: Maseno University, Kisumu, Kenya
Active Member Count: 4
Allocation Start: 2021-03-16
Allocation End: 2021-09-23
Used Hours: 85717
Project Name: Computational Study of Hematite Clusters
Project Shortname: MATS0888
Discipline Name: Material Science
Sunlight is the most abundant renewable energy resource and considered to be the ultimate solution to address the global energy problem: The Tetrawatt Challenge. However, the vision of solar
energy providing a substantial fraction of global energy infrastructure is still far from being realized. The major challenge is to develop an efficient and cost-effective approach for storing
solar energy that can be used on demand on a global scale. Splitting water photoelectrochemically is a promising way in which solar energy can be harvested and stored through
photocurrent-driven chemical bond rearrangement leading to the formation of molecular hydrogen (2H2O→2H2+O2, E0=1.23 eV, Standard Hydrogen Electrode, SHE), which can be used directly as fuel or as
an intermediate in the production of other fuels.
Hematite (alpha-Fe2O3) has emerged as a candidate photoanode material due to several favorable features: a band gap of about 2.0 eV (which allows for a large fraction of solar spectrum to be
absorbed), excellent stability under aqueous environments, and abundance in nature. Converting energy from the Sun to electricity efficiently via photoelectrochemical splitting of water is based on photo-induced charge separation at an interface (semiconductor/water interface) and the mobility of photogenerated carriers (electron-hole pairs). However, in bulk hematite, the separated charges recombine very fast (very short lifetimes of electron-hole pairs) and have very low mobility, hindering the conversion efficieny. These limitations can be dealt with by nanostructuring hematite, by growing ultrathin films of hematite, and by introducing impurity atoms i hematite. In this work, the structure and energetics of various size clusters of Fe2O3 have been investigated using the plane-wave pseudopotential approach to Density Functional Theory (DFT) and ab initio Molecular Dynamics (MD).
These calculations require a significant amount of compute resources in terms of memory and cpu scaling. In this respect, I am grateful to the CHPC for granting me access to their supercomputing facilities and I hope for more collaboration in future.
Principal Investigator: Dr Aniekan Ukpong
Institution Name: University of KwaZulu-Natal
Active Member Count: 3
Allocation Start: 2021-03-16
Allocation End: 2021-09-12
Used Hours: 88094
Project Name: Theoretical and Computational Condensed Matter and Materials Physics
Project Shortname: MATS0941
Discipline Name: Physics
Our research group has been working to provide a better understanding of materials, their interfaces, and the interactions they undergo. Our computational studies also lay the foundation for the technological application of materials based on the discovery of new physics. To this extent, first-principles ground-state calculations were combined with the dynamic evolution of a classical spin Hamiltonian to study the metamagnetic transitions associated with the field dependence of magnetic properties in frustrated van der Waals ferromagnets. We have observed dynamically stabilized spin textures relative to the direction of spin quantization as stochastic solutions of the Landau–Lifshitz–Gilbert–Slonczewski equation under the flow of the spin current. We also investigated the spin signatures that arise from geometrical frustrations at interfaces and observed the emergence of a magnetic skyrmion spin texture and its formation under competing internal fields is characterized. We also performed the analysis of coercivity and magnetic hysteresis to reveal a dynamic switch from a soft to the hard magnetic configuration when considering the spin Hall effect on the skyrmion. We have also found that heavy metals in capped multilayer heterostructure stacks host field-tunable spiral skyrmions. These skyrmions are topologically protected objects that could serve as unique channels for carrier transport. We have also demonstrated that dynamically switchable magnetic bits made from spiral skyrmions can be used to read and write data without the need for a spin-transfer torque. These results are important in the spin transport engineering of spintronic devices.
Principal Investigator: Mr Steven James
Institution Name: University of the Witwatersrand
Active Member Count: 5
Allocation Start: 2021-03-16
Allocation End: 2021-10-14
Used Hours: 195041
Project Name: Learning abstract representations for high-level planning
Project Shortname: CSCI1368
Discipline Name: Computer Science
We are a research group working in artificial intelligence as part of the RAIL Lab at the University of the Witwatersrand. Our work looks at how robots can use their sensory observations to learn symbolic concepts of their environment in a way that is similar to humans. For example, when planning, humans reason about the world using higher-order concepts, which allow us to ignore the unimportant details and construct long-term plans. The ability to plan far into the future is a key aspect of human intelligence and is something that must be achieved if we are to deploy robots in the real world.
Our approach relies on the robot being equipped with a set of skills, and then autonomously learning high-level concepts to support planning with those skills. Our work allows these robots to transfer these concepts to new unseen environments, which is an important step towards creating flexible robots capable of solving a variety of tasks. Ultimately, we envision our approach being used to create robots capable of operating in multiple environments, such as kitchens, hospitals and warehouses, reducing the need to develop costly, single-purpose robots.
Our approach relies on the CHPC's cluster to run numerous experiments with varying amounts of data to measure how efficiently we can construct these higher-order concepts. These results are further averaged over many trials to improve the statistical significance of our results and ensure that the outcome was not simply a fluke. We have currently reached a stage where the first phase of the project is complete - experiments in simulated environments have validated our approach. Our next step is to deploy our work on a real, physical robot to demonstrate its real-world applicability.
Principal Investigator: Dr Aijaz Ahmad
Institution Name: University of the Witwatersrand
Active Member Count: 1
Allocation Start: 2021-03-18
Allocation End: 2021-09-14
Used Hours: 13124
Project Name: Molecular modelling and drug design
Project Shortname: HEAL1257
Discipline Name: Health Sciences
We are exploring the anti-microbial and anti-viral drugs by targeting various enzymes. We are implementing virtual screening, pharmacophore modeling and molecular dynamic simulation studies to discover these drugs as potential therapeutics in various infectious diseases.
Principal Investigator: Prof Rotimi Sadiku
Institution Name: Tshwane University of Technology
Active Member Count: 8
Allocation Start: 2021-03-18
Allocation End: 2021-09-14
Used Hours: 1166
Project Name: NANOCOMPOSITE FOR ENERGY STORAGE
Project Shortname: MATS1376
Discipline Name: Material Science
This research group is made up of masters and doctoral student across different fields of engineering and are supervised by Prof. Rotimi Sadiku, Prof. Suprakas S. Ray, Prof. Hamam, and Prof. Bonex.
The group is mandated with the development of a high-performance energy storage system. Therefore, to successfully design and develop such material, the research team needs the use of High-Performance Computing resources. This is necessary in order to simulate at the atomistic level and understand the transport properties of the metal ions in the anode and how these ions arrange within the anode during charging and discharging. Without having a proper understanding of the behavior of this material using computational studies, the experimental work will simply be costly.
As such, this research study will employ atomistic and multiscale modeling to investigate the ion transport, absorption energy, diffusion barrier kinetics for metal ion intercalation, the specific capacity, voltage profile, power density, charging, and discharging cycle of the material to be developed for high-performance storage.
Principal Investigator: Dr Donald Mkhonto
Institution Name: Tshwane University of Technology
Active Member Count: 2
Allocation Start: 2021-03-18
Allocation End: 2022-01-27
Used Hours: 31111
Project Name: Electronic Structure and Atomistic Simulations of Minerals and Mineral Surfaces
Project Shortname: MATS0916
Discipline Name: Physics
Materials Modelling is done at Tshwane University of Technology in the department of Physics with Dr. Donald Mkhonto as the Principal investigator. Employing computational materials science techniques to study both bulk structure and surfaces of materials and minerals in general; Computational Materials Science at the department is aimed at using simulation methods to predict properties of different materials at different conditions that may be difficult and expensive experimentally. The methods allow us to calculate and predict materials properties at different length scales (from Angstrom to Micrometer length, that is from few atoms to thousands of atoms) While computational codes and hardware used for the kind of study may be expensive to academics, the use of CHPC (Centre For High Performance Computing) as provided by the government of South Africa is of vital importance and assist invaluably in terms of both physical computer infrastructure and software. We are able to introduce students (both undergraduate and post graduate) to computational materials science as an alternative and partner to experimental work and materials design.
Principal Investigator: Prof Koop Lammertsma
Institution Name: University of Johannesburg
Active Member Count: 2
Allocation Start: 2021-03-18
Allocation End: 2021-09-14
Used Hours: 304547
Project Name: ChiralCat
Project Shortname: CHEM1410
Discipline Name: Chemistry
The ChiralCat project is conducted at the University of Johannesburg under the direction of Prof. K. Lammertsma with Prof. A. Muller as co-supervisor and Dr. G. Dhimba as computational chemist and with PhD and MSc students performing experimental studies.
ChiralCat is about chiral-at-metal catalysis in which the chiral integrity of the transition metal catalysts is maintained during a chemical reaction. The rational design of such catalysts capable of effecting enantioselective chemical transformations is of paramount importance to satisfy the increasing industrial demand for chiral fine chemicals. ChiralCat is an innovative approach in asymmetric catalysis, requiring a detailed understanding to advance the field. So far, asymmetric catalysis is dominated by catalysts carrying expensive chiral ligands. Such conventional catalysts require exhaustive screening of the chiral ligand pool to obtain products with high enantiomeric excess. Not only is this a tedious and costly process, also the chiral ligands are often far more expensive than the transition metals. ChiralCat explores instead the use of abundantly available transition metals with readily available simple ligands to compose catalysts that are inherently chiral and that keep their chiral integrity during the catalytic reaction.
To provide these insights and assist experimentalist in synthesizing chiral-at-metal catalysts requires insight in the molecular behavior of the catalysts and their catalytic reactions. Computational chemistry is by far the best and most effective means to provide this insight, which may well simplify many industrial processes. Because of the available and indispensable compute power of CHPC we could make much progress in the first six months of this project showing the feasibility of asymmetric epoxidation of olefins with a simple molybdenum catalyst.
Principal Investigator: Dr Gerhard Venter
Institution Name: University of Cape Town
Active Member Count: 6
Allocation Start: 2021-03-23
Allocation End: 2021-09-19
Used Hours: 985209
Project Name: Simulation of Ionic Liquids
Project Shortname: CHEM0791
Discipline Name: Chemistry
Dr Venter's research group is part of the Scientific Computing Research Unit at the University of Cape Town and uses computer simulation to study the properties of a special class of solvents, called "ionic liquids". A fundamental understanding of the physical properties of ionic liquids and transports properties of solutes in these systems can lead to the rational design of new, environmentally friendly liquids with applications as electrolytes in next-generation batteries and energetic materials. The simulations not only provide first-principles characterization of ionic liquids and solutes, but also form the basis of models that aim to use machine learning for the prediction of thermodynamic properties.
High performance computing is required for all research conducted in this group and without the resources and infrastructure provided by the Centre for High Performance Computing (CHPC), the research will not be possible.
Principal Investigator: Dr Mehdi Mehrabi
Institution Name: University of Pretoria
Active Member Count: 1
Allocation Start: 2021-03-20
Allocation End: 2021-09-16
Used Hours: 3312
Project Name: Modeling of heat transfer characteristics and pressure drop of nanofluids in micro and mini tubes.
Project Shortname: MECH1086
Discipline Name: Other
I am a member of Clean Energy Research Group at University of Pretoria and I am particularly interested in numerical simulation of fluid flow and heat transfer for different application. My research group are focused at the moment to link the numerical simulation and AI to be able to decrease the time of the simulations and we are at the beginning of a long road. The progress of the project is good so far, in 2021 three high level journal articles were published in my research group and three are two manuscript under preparation.
Principal Investigator: Mr Adebayo Azeez Adeniyi
Institution Name: University of KwaZulu-Natal
Active Member Count: 6
Allocation Start: 2021-03-23
Allocation End: 2021-09-19
Used Hours: 29744
Project Name: Drug Discovery Research Using Quantum and Molecular Dynamic Simulation
Project Shortname: CHEM0958
Discipline Name: Chemistry
CHPC has significantly enhance our research impact and productivity this year which has already resulted into 9 published research articles. Our research focus has been expanded to include the vaccine development, photochemistry and polymer chemistry in addition to our area of research on drug development and electronic properties of small molecules. We make used of several packages that are installed on CHPC for Bioinformatics, Quantum and Molecular Dynamic Simulation. The reason for expanding our research focus is based on the pressing need of the society and global challenges, that lead us into epitopes based vaccine development from the genomes of the pathogens, polymer chemistry, photochemistry of small molecules for application as photoactive compounds and molecular electrochemistry. Our research use theoretical modelling to give rational insight into identified problems which can help in the experimental designs for possible solutions. CHPC is an important facility to our research group, we make use of the facilities and the softwares like Gaussian, Orca, Games and Newchem; also used molecular dynamic packages like Gromacs, Amber and Lammps. We have applied quantum packages to study the chemical and spectroscopic properties of small molecules of which we have been able to reproduced some of the experimental results.
It is obvious that our research progress and achievement revolve around the service of CHPC and many of our research output would not have been possible in without the support from CHPC.
Principal Investigator: Prof Rebecca Garland
Institution Name: University of Pretoria
Active Member Count: 6
Allocation Start: 2021-03-23
Allocation End: 2021-09-19
Used Hours: 285642
Project Name: Simulating atmospheric composition
Project Shortname: ERTH0846
Discipline Name: Earth Sciences
Air pollution can have large negative impacts on human health, agriculture, ecosystems, visibility and climate. In South Africa, although ambient air quality is regulated, many areas are out of compliance with the National Ambient Air Quality Standards. In order to protect human health and mitigate impacts, it is critical to improve air quality. The Constitution provides that everyone has a right to have an environment that is not harmful to their health. The Atmospheric Composition Focus Area (ACFA) in the CSIR Climate and Air Quality Modelling research group aims to provide the evidence base to quantify the impacts of air quality and to improve air quality.
The group uses the CHPC to run an air quality model to simulate urban and regional air quality at high resolution. This chemical transport model simulates the physics and chemistry of the atmosphere. The processes represented within the model are complex, and thus computationally intensive, which makes use of the CHPC facility a necessity. The CSIR ACFA is the only group in South Africa routinely running a chemical transport model to simulate air quality for management purposes.
Using the CHPC resources, the team has been able to simulate the impact of policy interventions on air quality in cities. Additionally, the team has simulated the health risk from air pollution regionally in South Africa. In the past, this has been done using monitoring station data only, which then limits the analysis to only those living directly around the station. The team has also simulated the impact of climate change on air pollution. These simulations use the climate projections from the CSIR's CCAM-CABLE, which is also run at CHPC, to provide meteorology input into the air quality model.
These outputs directly provide the evidence base needed for decision makers to draft and implement policies and interventions to effectively improve air quality as well as understand its impacts, now and into the future.
Principal Investigator: Prof Kevin Naidoo
Institution Name: University of Cape Town
Active Member Count: 12
Allocation Start: 2021-03-24
Allocation End: 2021-09-20
Used Hours: 739358
Project Name: Reaction Dynamics of Complex Systems
Project Shortname: CHEM0840
Discipline Name: Chemistry
The Scientific Computing Research Unit (SCRU) directed by Prof Kevin Naidoo is built from the following laboratories:
Computation & Modelling Laboratory
Cancer Translational Science Laboratory
Informatics & Visualisation Laboratory
All three laboratories make direct or indirect of CHPC resources to develop diagnostics for Cancer or Therapeutics in Biomedicine. Of national interest is the collaboration between the SCRU laboratories and the CHPC to make advance computation more easily accessible to South African researchers in the fields of Biology and Chemistry through the CHPC's hosting of the Galaxy Biomolecular Reaction Interaction Dynamics Global Environment (BRIDGE). Prof. Naidoo's research groups are mostly interested in the development and applications of methods useful to Life Scientists. He is a domain specialist in code development for Free Energy methods, Molecular Dynamics and unsupervised Machine Learning. Flagship codes developed by Prof Naidoo and his students are the Free Energy from Adaptive Reaction Coordinate Forces (FEARCF), Free Energy Force Induced (FEFI) coarse grained molecular dynamics and the Denoising Autoencoder Self Organising Map (DASOM). The SCRU Cancer Translational Science Laboratory links the computation and informatics technologies with laboratory experiments to provide models for cancer research scientists aiming to achieve Translational Research goals cancer care
Principal Investigator: Dr Adam Skelton
Institution Name: University of KwaZulu-Natal
Active Member Count: 9
Allocation Start: 2021-03-23
Allocation End: 2021-09-29
Used Hours: 296741
Project Name: Molecular modelling of biomolecules and materials
Project Shortname: CHEM0798
Discipline Name: Chemistry
I am Dr. Adam Skelton and I work at the College of Health Science university of KwaZulu-Natal. I work on understanding the behaviour of biological and material systems on the molecular level.
A particlar interest is in using molecular simulation to direct the rationale design of functionalized nanoparticles for a range of applications such as drug delivery, water treatment and catalysis. In particular, drug delivery is the use of nanoparticles to carry antiviral or antibacterial drugs into parts of the body that are difficult for conventional drugs to reach. Drug-delivery is, therefore, of huge importance in South Africa, especially in the battle against HIV and Tubercolosis.
The research requires running a series of computer simulations, which may take hours at a time. CHPC has had a great impact since it provides me with the facilities to perform such simulations.
Since the project started, my group has published more than 25 papers in international journals and has provided deep insight into biological and material systems on the molecular level.
Principal Investigator: Prof Graham Jackson
Institution Name: University of Cape Town
Active Member Count: 4
Allocation Start: 2021-03-24
Allocation End: 2021-09-20
Used Hours: 7285
Project Name: Insect neuropeptides
Project Shortname: CHEM1101
Discipline Name: Chemistry
Using intensive computational methods, and targeting insect neuropeptides, new, species specific, 'green', pesticides are being developed against the locust. These pesticides are designed to block the energy pathway of the insect.
Principal Investigator: Prof Randhir Rawatlal
Institution Name: University of KwaZulu-Natal
Active Member Count: 1
Allocation Start: 2021-03-25
Allocation End: 2021-09-21
Used Hours: 68951
Project Name: CFD modelling of methane oxidation in monolith reactors
Project Shortname: MECH0894
Discipline Name: Chemical Engineering
The CHPC is an important national resource which supports research and industry. In these days where all companies want to develop 4IR capacity, it really is a national treasure.
Principal Investigator: Dr Samuel Iwarere
Institution Name: University of Pretoria
Active Member Count: 2
Allocation Start: 2021-03-26
Allocation End: 2021-09-22
Used Hours: 257357
Project Name: Understanding Plasma-Liquid Interaction
Project Shortname: MECH1107
Discipline Name: Chemical Engineering
The research is being carried out under the guidance of the Thermodynamics Research Unit based in the Discipline of Chemical Engineering at the University of KwaZulu-Natal. The understanding of plasma-liquid interactions is of prime importance in the context of environmental remediation and wastewater treatment applications. Plasma treatment of wastewater involves complex chemical and physical phenomena, which makes the comprehensive modelling of phenomena such as gas phase reaction, turbulence and heat transfer a difficult task. In order to gain insight into the complex processes associated with the use of plasma wastewater treatment, Computational Fluid Dynamics (CFD) will be used to study the relevant chemistry and fluid dynamic effects at the plasma-liquid interface. The Centre for High Performance Computing (CHPC) allows for calculation within realistic time frames, allowing the researchers to determine the accuracy of the simulations encountered in plasma physics. Simulation of the plasma arc is progressing well with the researchers working in conjunction with other scientists in the plasma field.
Principal Investigator: Prof Sybrand van der Spuy
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-03-26
Allocation End: 2021-09-22
Used Hours: 614453
Project Name: CFD simulation of turbomachinery
Project Shortname: MECH1187
Discipline Name: Computational Mechanics
The turbomachinery research group at Stellenbosch University has been making use of the CHPC for their CFD simulation work during the past three years. The use of the CHPC enables the accurate and detailed modelling of aspects like gas turbine combustion and fan noise. The results achieved to date have been world class.
Principal Investigator: Dr Quinn Reynolds
Institution Name: Mintek
Active Member Count: 7
Allocation Start: 2021-03-29
Allocation End: 2021-09-25
Used Hours: 127927
Project Name: Computational Modelling of Furnace Phenomena
Project Shortname: MINTEK776
Discipline Name: Applied and Computational Mathematics
The study of pyrometallurgy, the engineering discipline focused on the use of high temperatures to process raw materials into valuable commodities such as iron and steel, precious metals, and many other industrially- and economically-significant products, is fraught with many challenges. Pyrometallurgical smelters are fierce environments hazardous to man and machine alike, and as a result our knowledge of what happens inside them is often severely limited.
One of the most significant problems faced in pyrometallurgy is the staggering range of scales at play. A high-temperature smelting process typically starts at the nanoscale, with chemical reactions strongly influenced by surface properties and complex thermodynamics of the materials involved. Moving upward to the micro-scale, interfacial film phenomena can dominate phase separation problems and cause significant operational problems if undesirable phase mixing occurs. Further up in scale to millimetres and centimetres and we enter the realm of particle mechanics, where millions or billions of individual particles fed to a furnace can interact to produce unexpected and counter-intuitive granular flow structures. At the scale of metres we have the furnace vessels themselves, containing numerous coupled multiphysics systems. Finally at the scale of kilometres we have the environment surrounding the smelter plant, all of which can be influenced (positively or negatively) by its presence.
Engineers and scientists at Mintek rely on theoretical knowledge of the processes involved at all scales to create detailed mathematical descriptions which can be solved using powerful computers such as those available at the Centre for High Performance Computing. Such digital experiments give us unprecedented insight into the inner workings of pyrometallurgical furnaces, and let us advance the state of the art by improving their energy efficiency, reducing their environmental impact, and developing new process routes for complex feed materials such as recycled wastes.
Principal Investigator: Dr Anna Bosman
Institution Name: University of Pretoria
Active Member Count: 12
Allocation Start: 2021-03-29
Allocation End: 2021-09-25
Used Hours: 101774
Project Name: Deep Learning and Neural Network Research at CS UP
Project Shortname: CSCI1166
Discipline Name: Computer Science
The research program of the Computer Science Department, University of Pretoria (UP), is headed by Dr Anna Bosman, and forms a part of the Computational Intelligence Research Group (CIRG) at UP. The focus of the program is on the following research topics: fitness landscape analysis of neural networks and real-life applications of deep learning. The following deep learning applications are currently investigated: (1) convolutional neural networks for satellite image analysis; (2) evolution of recurrent network architectures; (3) convolutional neural networks for image segmentation; (4) convolutional neural networks applied in the entomology domain.
Studying fitness landscapes of neural networks deepens our general understanding of the fundamental principles of neural network research, and enables progress in fundamental artificial intelligence. Deep learning applications allow us to apply modern artificial intelligence techniques in the South African context. Thus, the usage of CHPC helps us strengthen machine learning and artificial intelligence research in South Africa.
Sampling and estimation techniques for fitness landscape analysis are used to probe the error surfaces of the deep and shallow neural networks. The research is of empirical nature, and entails heavy sampling of the search space. Such experiments would not be feasible without access to a computer cluster. Deep learning applications benefit significantly from being run on a GPU, therefore, access to the GPUs allows us to upscale the experiments to real-life applications. The research program yields yearly publications in highly ranked international journals and conferences. This indicates that the research program is productive, and successfully delivers academic outputs.
Principal Investigator: Prof Juliet Hermes
Institution Name: SAEON
Active Member Count: 6
Allocation Start: 2021-03-29
Allocation End: 2021-09-25
Used Hours: 74786
Project Name: SAEON Coastal and Regional Ocean Modelling Programme
Project Shortname: ERTH1103
Discipline Name: Earth Sciences
The Coastal and Regional Ocean Modelling Programme is a modelling initiative of the South African Environmental Observation Network (SAEON) Egagasini Node based in Cape Town. The aim of this programme is to use ocean models to understand regional and coastal shelf dynamics. While global models exist, and perform relatively well at the basin-scale, in order to capture the important, fine scale and highly variable processes of our shelf and coastal regions, we are developing downscaling approaches. These use the outputs from global models to force high resolution regional models that can be coupled with wave, atmospheric or biogeochemistry models.
Principal Investigator: Dr Angeline Wairegi
Institution Name: Strathmore University, Nairobi, Kenya
Active Member Count: 1
Allocation Start: 2021-03-29
Allocation End: 2021-09-25
Used Hours: 7767
Project Name: Investigating AI in Africa
Project Shortname: CSCI1412
Discipline Name: Other
The research group is based at the Center of Intellectual Property and Information Technology, a think tank and policy research center based in Nairobi, Kenya, and housed under Strathmore University. The research group's overarching goal is to investigate the AI ecosystem on the continent. There is minimal research that is looking in the development, adoption, governance and impact of AI in Africa. The African voice has largely been excluded in the global conversations around AI - we seek to change that. Specific to the use of the HPC platform, we seek to use it to ascertain: (i) the best way of optimizing the racial ratios of open source facial datasets to minimize bias in their use in the African continent; (ii) develop an AI platform that can be used to identify localized potato disease, and (iii) determine whether AI may be used in a predictive capacity, i.e. predicting case rulings given a set of parameters, in the labor courts in the Kenyan judiciary. Each of these 3 projects require the use of large datasets to train an algorithm; the HPC, with its high computational power, cuts down on computational time. The facial recognition project is nearing completion while the other 2 projects are in their early phases.
Principal Investigator: Dr Caroline Kwawu
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 13
Allocation Start: 2021-03-31
Allocation End: 2021-11-11
Used Hours: 2244915
Project Name: Renewable Energy Materials
Project Shortname: MATS1146
Discipline Name: Material Science
Renewable energy materials is a research being undertaken by students in the Department of Chemistry, Kwame Nkrumah University of Science and Technology. The Centre for High Performance Computing (CHPC), South Africa, provides computing resources and access to software that makes the research possible. The aim of the research is to model new materials that convert solar energy to fuels and electricity. This work is peculiar and urgent to reduce the over reliance on fossil fuels which pose the problem of Climate Change. This computational group also collaborates with an experimental group to understand their findings and to better tune materials by doping and surface alterations. We employ the Density Functional Theory (quantum) method which is efficient in studying electronic structure to elucidate surface reaction mechanisms, and properties of materials. This research is important to the migration of Ghana from a carbon dioxide emitting economy to a mitigating economy. As Ghana currently produces much green house gases from biomass combustion, due to its cost. This allows Ghana to also contribute to the SDGs (Sustainable Development Goals) of finding renewable, sustainable energy for all.
Principal Investigator: Dr Lucy Kiruri
Institution Name: Kenyatta University, Nairobi, Kenya
Active Member Count: 7
Allocation Start: 2021-03-31
Allocation End: 2021-09-27
Used Hours: 255786
Project Name: KU Computational Chemistry Research Group
Project Shortname: CHEM1032
Discipline Name: Chemistry
The members of the group are in two different areas of Molecular modeling, that is; quantum mechanical calculations and Molecular Dynamics. For the QM calculations, we largely use the Gaussian 16 program and for MD simulations we use the GROMACS 2020. We have a new member in the group (Dr. Abdulhameed, Adamu) who is working on his post-doctoral research. His work will involve the adsorption of dyes and /drugs on chitosan. He will use Gaussian 16 and also Gromacs to accomplish the study.
The other member of the group is my collaborator, Dr. Daniel Shadrack, who is a Lecturer at St. John University Dodoma. Daniel's work is on MD simulations of natural products targeting important proteins for Covid-19 treatment. In addition, we are collaborating on research involving molecules that have cation-pi interactions and pi-pi interactions. Our first paper on this work has been accepted and is under the proofreading section.
Principal Investigator: Mr Jean Pitot
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-03-31
Allocation End: 2021-09-27
Used Hours: 94207
Project Name: SAFFIRE Programme
Project Shortname: MECH1121
Discipline Name: Computational Mechanics
The University of KwaZulu-Natal's Aerospace Systems Research Group (ASReG) is at the forefront of rocket propulsion research in academia, both in South Africa and Africa as a whole. Located at the University's Discipline of Mechanical Engineering, ASReG primarily conducts research in the areas of liquid and hybrid rocket propulsion, as well sounding rocket development. The SAFFIRE Programme aims to develop South Africa's first fully-integrated liquid rocket engine, which is being designed for use in a small satellite launch vehicle. The long-term intention of the programme is for the engine to enable the indigenous development of a South African space launch capability. To meet its developmental goals, the programme requires the simulation of complex computational models to predict the fluid dynamic, thermochemical and structural performance of various engine components. These simulations are typically very computationally-expensive, and require extensive computational resources to run within a reasonable time frame. The CHPC offers the only realistic means of obtaining access to such resources. The programme is progressing well, and 2021 has seen the intake of numerous new postgraduate and undergraduate students. In total, 16 postgraduate students are currently involved in ASReG's cutting-edge projects.
Principal Investigator: Dr Zipporah Muthui
Institution Name: Chuka University, Chuka, Kenya
Active Member Count: 7
Allocation Start: 2021-04-01
Allocation End: 2021-09-28
Used Hours: 2380
Project Name: Electronic Structure and Magnetic Properties of Heusler Compounds
Project Shortname: MATS1112
Discipline Name: Physics
MATS1112: Electronic Structure and Magnetic Properties of Heusler Compounds research group is composed of the PI, Dr Zipporah Muthui and students Ms.Jane Mbae, Ms. Eunice Gitonga, Mr Bonnface Mwanzia, Mr Gabriel Mutava and Mr Robinson Owino, all from Chuka University. We are mainly involved in studying the electronic structure of materials for various applications including spintronic, photocatalytic, piezzo electric and 2D materials. The systems studied will involve large supercells that require their calculations to be performed on a high performance computing facility, as single computers would not be sufficient to handle the systems. The project is progressing well. and with increased knowledge, number of students and software, the research output is expected to increase.
Principal Investigator: Prof Arnaud Malan
Institution Name: University of Cape Town
Active Member Count: 6
Allocation Start: 2021-04-01
Allocation End: 2021-09-28
Used Hours: 98583
Project Name: UCT CFD
Project Shortname: MECH1314
Discipline Name: Computational Mechanics
InCFD is a research group based within the department of Mechanical Engineering at the University of Cape Town (UCT). The prime objectives are to develop state-of-the-art computational fluid dynamics (CFD) modelling tools and CFD based innovation for the express support of industry. This is done via the uncompromising pursuit of innovation through fundamental research which carries the hall mark of excellence. InCFD is home to the South African Research (SARChI) Chair in Industrial CFD. Between April and Sept we have been working on refining our multi-phase slosh modelling capability and doing rigorous comparisons to experimental data for the EU project SLOWD i.e. comparing to experimental data with 98% accuracy demonstrated for violent slosh forces.
Principal Investigator: Dr Richard Walls
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-04-06
Allocation End: 2021-10-03
Used Hours: 97720
Project Name: Fire Engineering Research
Project Shortname: MECH1148
Discipline Name: Computational Mechanics
Ongoing work has sought to understand fire testing furnaces such that everything from doors to communication cables can be tested in a consistent manner. This work is providing an understanding of how the overall process can be made more efficient and consistent.
Principal Investigator: Dr Calford Otieno
Institution Name: Kisii University, Kisii, Kenya
Active Member Count: 6
Allocation Start: 2021-04-06
Allocation End: 2021-10-03
Used Hours: 1980
Project Name: FIRST PRINCIPLE ELECTRONIC STRUCTURE CALCULATIONS OF EMERGING MATERIALS FOR SOLAR CELL APPLICATIONS.
Project Shortname: MATS1083
Discipline Name: Physics
My research group identifies as the Computational Material and Theoretical group(C MTG), We are based at Kisii University-Main Campus, in Kisii town in Kenya. To process our data and to perform many complex calculations we have always used the resources at CHPC to Implement our Density Functional Theory scheme. Our groups essentially study FIRST PRINCIPLE ELECTRONIC STRUCTURE CALCULATIONS OF EMERGING MATERIALS FOR SOLAR CELL APPLICATIONS via Density Functional Theory using the open-source software Quantum Espresso. Currently, our main focus is on the First Principle Study of Electronic, Structural and Mechanical Properties of Iron pnictides because of their promise as materials of the future. We investigate the effects of doping on bandgap engineering since bandgap Physics is key to energy conversion and material excitations. We also investigate the effect of pressure with an aim to improving these materials for their future industrial applications. Finally, all the progress we have made is courtesy of the CHPC
Principal Investigator: Dr Luna Pellegri
Institution Name: University of the Witwatersrand
Active Member Count: 1
Allocation Start: 2021-04-06
Allocation End: 2021-10-21
Used Hours: 122130
Project Name: Developing of new techniques for real-time verification in proton-therapy
Project Shortname: PHYS1327
Discipline Name: Physics
GEANT4 simulations were performed to study the effect of selected elements in the prompt gamma emission during proton therapy. The results obtained are compared with experimental measurements to investigate the validity of the simulations and the quality of the element selected.
Principal Investigator: Prof Ponnadurai Ramasami
Institution Name: University of Mauritius
Active Member Count: 5
Allocation Start: 2021-04-08
Allocation End: 2021-10-14
Used Hours: 195564
Project Name: Computational Chemistry Methods to Study Structural and Spectroscopic Parameters
Project Shortname: CHEM1290
Discipline Name: Chemistry
I am the group leader of the Computational Chemistry Group in the Department of Chemistry of the University of Mauritius. The research interest of the group ranges from fundamental to applied chemistry. One of the research interests is to predict structural and spectroscopic parameters of novel compounds. In this context, we are currently predicting the structures of two-dimensional nanomaterials prior to their synthesis. We are also studying the electronic structure of metal oxide nanoparticles. These projects involve the use of Quantum Espresso.
We are also collaborating with experimentalists, mainly organic and inorganic chemists from African countries. From the organic chemistry perspective, we are studying cycloaddition reactions to provide insights into the reaction mechanism so as to explain the formation of the observed products. From the inorganic chemistry aspects, we are studying inorganic complexes using computational method so as to complement the experimental research in terms of structural and spectroscopic parameters. However, in some cases, where experimental data are not available, the computed parameters are useful to the experimentalists for the interpretation of results. For these projects, we are using Gaussian software.
Without the use of the CHPC facility, it would not have been possible to carry out high level computations for the projects that we have planned. This facility is allowing us to aim at carrying out high level research and to eventually have good publications in high impact factor journals.
We have recently started using the CHPC facility and we are currently still running calculations and interpreting data. We have one publication and we have several drafts which we will submit for publications soon.
Principal Investigator: Dr Winfred Mulwa
Institution Name: Egerton University, Egerton, Kenya
Active Member Count: 9
Allocation Start: 2021-04-08
Allocation End: 2021-10-14
Used Hours: 339325
Project Name: Magnetic Refrigeration
Project Shortname: MATS1181
Discipline Name: Physics
Who: MATS1181 under PI Winfred Mulwa from Egerton University Kenya: Magnetic refrigeration. What: Magnetization and demagnetization of magnetic materials known as magnetic refrigeration that rely on magnetocaloric effect (MCE). In cooling technology, magnetic refrigeration which depends on magnetocaloric effect is commonly used in refrigerators to achieve exceptionally low temperatures. Why: . Magnetic refrigeration does not rely on the uses of harmful and ozone-depleting coolant gases. How: This work is done computationally. We use the Quantum Espresso code (Density Functional Theory). This purely depends on CHPC because all the calculations have to be done in CHPC. Thanks to CHPC. How is the project progressing: We have achieved all the objectives. The project is as planned. Two of my MSc students are graduating in December 2021
Principal Investigator: Dr Sphelele Sosibo
Institution Name: North-West University
Active Member Count: 5
Allocation Start: 2021-04-08
Allocation End: 2021-10-05
Used Hours: 1188
Project Name: Molecular Dynamics of target enzymes
Project Shortname: HEAL1414
Discipline Name: Chemistry
The Organic and Medicinal Research Group uses the CHPC to identify the interaction of plant based photochemical compounds against various protein disease targets.
The isolation of plant bioactive material is a naturally destructive process. Computational Chemistry allows for studying of chemical entities without the use of actual chemicals. This reduces the number of experiment that can be performed in in-vitro and in-vivo testing.
We are using the CHPC parallel computing facilities to scale some of the project we carry in personal computers. We have had success in publishing some of the work from CHPC.
There is currently a good perception of the CHPC services from the students and it is envisioned that most of the work will be published.
Principal Investigator: Dr Vuyani Moses
Institution Name: Rhodes University
Active Member Count: 28
Allocation Start: 2021-04-08
Allocation End: 2021-10-05
Used Hours: 2114974
Project Name: Bioinformatics and Computational Chemistry Applications
Project Shortname: CBBI1122
Discipline Name: Bioinformatics
Our research group which falls under the Rhodes university research Unit in Bioinformatics (Rubi). Our interests are using computational techniques to solve biological problems such as drug discovery and biofuel production. we cater to prost graduate students from Hon, MSc and PhD.
Principal Investigator: Prof Rosemary Dorrington
Institution Name: Rhodes University
Active Member Count: 8
Allocation Start: 2021-04-08
Allocation End: 2021-10-14
Used Hours: 53549
Project Name: Marine Natural Products Research
Project Shortname: CBBI0963
Discipline Name: Bioinformatics
Marine Natural Products Research Group, Department of Biochemistry and Microbiology, Rhodes University
The Marine Natural Products Research Platform covers the broad field of marine biodiscovery, focusing on the potential of bioactive secondary metabolites produced by marine organisms endemic to the Agulhas Bioregion as lead compounds for drug discovery.
Overview and aims
More than half of new pharmaceutical drugs on the market or in clinical trials are natural products and their derivatives and the majority of these are now coming from marine macrofauna and their associated microbiomes. The Maine Natural Produces research programme is active across the broad field of marine biodiscovery, focusing on exploring the potential of marine invertebrates and their associated microorganisms as sources of novel bioactive small molecules. Research projects include: (1) Biodiversity mapping of benthic habitats in the Agulhas bioregion, focusing on invertebrates, ascidians, soft corals and their associated microbiota; (2) Isolation and characterization of bioactive secondary metabolites of marine organisms; (3) screening marine natural product libraries for anti-viral, anti-cancer and antibacterial activity and (4) elucidating the metabolic pathways that result in the production of lead compounds to pave the way for engineering recombinant systems for their production for the pharmaceutical industry. Associated projects focus on the application of high throughput metagenomic and analytical chemistry technologies to map patterns and processes in marine ecosystems to provide tools for sustainable development of our ocean economy. All of these research activities require access to high performance computing facilities to curate, process and analyse large Next Generation Sequencing (NGS) datasets. Primarily, we have used the CHPC facilities to analyse NGS amplicon library data to characterise the microbial communities of both marine and terrestrial systems.
Principal Investigator: Prof Eno Ebenso
Institution Name: North-West University
Active Member Count: 6
Allocation Start: 2021-04-08
Allocation End: 2021-10-05
Used Hours: 879336
Project Name: Theoretical Investigations of Corrosion Inhibition Activities of Organic Compounds
Project Shortname: CHEM1176
Discipline Name: Chemistry
Our research activities centre on computational materials science, which involves the design of new materials with improved performances and properties. These include the design of corrosion and scale inhibitors, chemical and biological sensing materials etc. We have started looking at the design of photoactive materials in the recent time. As we diversify research activities, we have included computational environmental remediation and reaction mechanism monitoring. The group utilizes computational chemistry software such as Gaussian and Materials Studio. Computational studies conducted in the group are mostly used to corroborate experimental results to ensure comprehensive scientific findings. Our continuous access to CHPC facilities has assisted us in timely execution of research plan and results acquisition. We have been able to achieve significant improvement in calculation speed when compared to our previous experience using PC.
Principal Investigator: Dr Matshawandile Tukulula
Institution Name: University of KwaZulu-Natal
Active Member Count: 1
Allocation Start: 2021-04-08
Allocation End: 2021-10-05
Used Hours: 3972
Project Name: Medicinal Chemistry and Computer-Aided Drug
Project Shortname: HEAL1346
Discipline Name: Chemistry
My research group at the University of KwaZulu Natal is part of the broader Natural Product Research Group (NPRG). Specifically, my group looks into the medicinal chemistry of infectious diseases of poverty related diseases such as Malaria and TB. We are looking at designing and evaluating bioactive molecules for the said diseases. The CHPC infrastructure allows us to undertake homology modelling, docking and chemoinformatic work that supports our synthetic endeavours. To date, we have made significant progress on Malaria work using the CHPC infrastructure.
Principal Investigator: Prof Stefan Ferreira
Institution Name: North-West University
Active Member Count: 5
Allocation Start: 2021-04-09
Allocation End: 2021-10-21
Used Hours: 187551
Project Name: Astrospheres
Project Shortname: ASTRO1277
Discipline Name: Astrophysics
The aim of this project is to successfully simulate astrospheres via a full three-dimensional magneto-hydrodynamic numerical model in order to compare results with telescope observations to gain a deeper understanding of these. From telescope observations there is very limited certainty on different evolution parameters like mass-loss rate, ambient density, magnetic fields, radiative cooling etc. The combination of such a code with observations will enhance our understanding of these cavities around stars. Also of importance is to calculate the cosmic ray distribution inside these cavities to account for their possible contribution to possible habitable planets close to the stars. For this purpose researchers from the North-West University and Center for High Performance Computing (CHPC) in South Africa has teamed up with two German Universities, namely the Ruhr-Universitat Bochum and the Christian Albrechts University in Kiel to bring together expertise. Central to this project is the successful implementation and execution of the Cronos MHD model at the CHPC. Currently we have successfully tested and obtained first results which will be used in a forthcoming publication.
Principal Investigator: Dr Ryno Laubscher
Institution Name: Stellenbosch University
Active Member Count: 4
Allocation Start: 2021-04-09
Allocation End: 2021-10-06
Used Hours: 377834
Project Name: University of Cape Town Applied Thermal-Fluid Research Unit
Project Shortname: MECH1279
Discipline Name: Computational Mechanics
Rising energy demand and the imminent threat of climate change are critical issues in society today. Thermofluid systems provide the backbone of almost all energy conversion processes for renewable and conventional power generation, as well as heating and cooling systems such as heat pumps and refrigeration cycles. The Applied Thermofluid Process Modelling Research Unit (ATProM) at the University of Cape Town specialise in modelling these systems to evaluate novel technologies, improve the efficiency and control of processes, and detect anomalies for condition monitoring purposes.
Fundamental models are built using detail Computational Fluid Dynamics (CFD) and integrated one-dimensional thermofluid networks. These are often combined with advanced optimization techniques and data-driven models that are derived via artificial intelligence (AI) and machine learning (ML) techniques. This unique combination of fundamental thermofluid principles and AI techniques enable the development of versatile and accurate numerical tools to address industry needs. However, applying these techniques often involve resource intensive computational procedures. Given this, the services and facilities provided by CHPC are invaluable in generating industry-scale research outputs.
The current focus of research is on supercritical CO2 (sCO2) Concentrated Solar Power (CSP) plants, optimization of heat pump systems, flexibility of fossil fired power plants (coal, gas turbines and combined cycles), energy storage systems, biomass energy conversion and Physics Informed Neural Networks (PINN) applied to energy systems.
Principal Investigator: Prof Chris Meyer
Institution Name: Stellenbosch University
Active Member Count: 4
Allocation Start: 2021-04-12
Allocation End: 2021-10-09
Used Hours: 236200
Project Name: Development of CFD models
Project Shortname: MECH1077
Discipline Name: Computational Mechanics
The research activity is led by Prof CJ Meyer from Stellenbosch University and is focused on developing computational models of industrial thermo- and fluid dynamic processes. There is a specific emphasis on the use of free software under the GNU General Public License. The industrial processes under consideration are those relevant to specific segments of industry in South Africa most notably power generation. Current projects include the simulation of large-scale air-cooled condensers as well as detailed simulations concerning the flow through axial flow fans used in the energy sector. The cohort of students are making excellent progress.
Principal Investigator: Prof Fourie Joubert
Institution Name: University of Pretoria
Active Member Count: 33
Allocation Start: 2021-04-12
Allocation End: 2021-10-14
Used Hours: 20678
Project Name: NGS Bioinformatics
Project Shortname: CBBI0905
Discipline Name: Bioinformatics
The field of bioinformatics (the application of mathematical, statistical and computing techniques on biological datasets) requires extensive computing resources. Reported here is research which focus on which cells in the body are beneficial in fighting HIV infections and new T-cell sub-populations have been identified. In the tick project, differences between ticks are studied together with the level of expression of proteins in ticks, for the production of better anti-tick vaccines. In the maize project, automated machine learning models are being developed, so that farmers can take photographs of maize leaves to automatically identify the infection of their crops by grey leaf spot, a common fungal infection of maize.
Principal Investigator: Dr Geoff Nitschke
Institution Name: University of Cape Town
Active Member Count: 15
Allocation Start: 2021-04-13
Allocation End: 2021-10-10
Used Hours: 969046
Project Name: Evolving Complexity
Project Shortname: CSCI1142
Discipline Name: Computer Science
This project is investigating novel machine-learning techiques for transfer learning - this is the transfer of skills learnt in one task to another more complex task. For now, simple virtual game-playing agents are being used, but the research aims to develop machine-learning techniques that can be applied to robotic control, autononmous vechicles and the like.
Principal Investigator: Dr Fabio Cinti
Institution Name: NITHEP
Active Member Count: 4
Allocation Start: 2021-04-14
Allocation End: 2021-10-11
Used Hours: 194932
Project Name: Quantum Monte Carlo for ultra-cold atoms
Project Shortname: PHYS0892
Discipline Name: Physics
A group led by Fabio Cinti (associate professor at University of Johannesburg) explores pattern formation of special structures like stripe phases, cluster crystals and quasicrystals in a quantum regime. Presently these structures can understand a large amount of fascinating phenomena in soft matter, superconductivity, nonlinear optical systems, .and long-range interacting systems in general. In this context quasicrystals are one of the most intriguing examples, as particles self-assemble in a long-range ordered pattern which is at the same time non-periodic, thus been able to exhibit forbidden crystalline ordering. Many studies observed cluster quasicrystals in soft macromolecular systems at finite temperatures by using this type of interactions. While a recent theoretical work have surprisingly revealed the stability of those structures also at zero temperature for a particular case, the extent to which classical cluster quasicrystals can be stable in the absence of thermal fluctuations is a matter of debate.
We investigate quantum cluster quasicrystals by imposing external quasi-periodic potentials to bosonic systems, so creating quasicrystalline structures in two-dimensional optical lattices. Interestingly, the competition of interactions and quasiperiodicity generate a wide range of phases, such as supersolidity and Bose glasses. We also observe superfluidity in a model relevant to quantum cluster quasicrystal. By using quantum Monte Carlo approaches, it was found that moderate quantum fluctuations make dodecagonal structures to persist, leading to a small but finite local superfluid phase.
By increasing fluctuations, a structural transition from quasicrystal to cluster triangular crystal takes place. Our studies bring to the conclusion that, at a quantum level, quasicrystal phases are produced as a joined effect of quantum fluctuations and a properly designed interaction potential between particles. This research have been carried out using the computational resources of the CHPC facilities.
Principal Investigator: Dr Lawrence Borquaye
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 5
Allocation Start: 2021-04-14
Allocation End: 2021-10-11
Used Hours: 256451
Project Name: Biomolecular Computations
Project Shortname: HEAL1382
Discipline Name: Chemistry
This is the Borquaye Research Group in the Department of Chemistry, Kwame Nkrumah University of Science and Technology, Kumasi-Ghana. Our research group focuses on bioactive molecules such as natural products, peptides, essential oils and/or small molecules and their biological activities. We also make use of molecular docking and molecular dynamics tools to unravel potential modes of action of bioactive compounds, predict their preferred binding targets and explore events that mediate biomolecule-ligand interaction over a specified time period. The goal is to unearth compounds that could potentially be developed into drugs and to understand how these compounds work in biological systems. Our molecular docking and molecular dynamics projects require high performance computing, which we are able to access at the CHPC (Lengau). We have made significant progress and have submitted our first set of manuscript for publication based on data derived from the Lengau. We also have different projects in various levels of completion. We have been able to get a clear understanding of how certain natural products and peptides interact with their specific protein targets and biological membranes.
Principal Investigator: Prof Emile Rugamika CHIMUSA
Institution Name: University of Kinsasha
Active Member Count: 12
Allocation Start: 2021-04-19
Allocation End: 2021-10-16
Used Hours: 185125
Project Name: Computational and statistical methodologies for human and environmental health prediction
Project Shortname: CBBI0818
Discipline Name: Bioinformatics
Computational and statistical methodologies for human and environmental health prediction research programme (CBBI0818) is running from the University of Cape Town, South Africa. This platform is fundamentally aiming 1) to determinate the environment and genetic variation that cause human species to look different, having difference in allergy, drug responses and treatment. 2) to identify and quantify the pattern of inheritance of the observed trait, drug response and treatment variation among human. These are addressed through the design of machine intelligence and artificial intelligent methodologies and statistical approaches to analysis large-scale DNA data of thousands affected/unaffected subjects. In doing so, this programme allows us leveraging large-scale and computational cost project to contributing to human health by increasing understanding of the genetic and environmental underpinnings of complex traits, drug/treatment responses and drug/dosage responses. Up to today, the use of parallel High-Performance Computing at CHPC has critically increased our researcher portfolio to meet the international standard with respect to large scale genomic era. This programme has already developed of genomic-based software tools that address African genetic variation challenges and provided advanced training around parallel programming with respect to large-scale genomic data to African postgraduate students/researchers.
Principal Investigator: Prof Marla Trindade
Institution Name: University of Western Cape
Active Member Count: 1
Allocation Start: 2021-04-19
Allocation End: 2021-10-16
Used Hours: 1058
Project Name: Microbial genomics
Project Shortname: CBBI1149
Discipline Name: Bioinformatics
Microorganisms carry a wealth of potential in their genomes. They can produce enzymes for industry and small molecules that could be antibiotics, anti-cancer drugs, anti-inflammatories and a host of other medically useful compounds as well as being used as whole cell biocatalysts. For humans to better understand how to best use this potential we need to unlock the genetic code in order to take advantage of what they can offer humanity. At the Institute for Microbial Biotechnology and Metagenomics (IMBM) we have isolated thousands of bacteria and their viruses from a variety of environments which we test for their ability to produce such proteins and compounds and determine how they function. One of our successes is the isolation of a novel ornithine lipid synthase from a oil contaminated site using a special technique called metagenomics. This gene allows for the production of a biosurfactant which are compounds which can replace chemically synthesized surfactant in future products. To get to grips with the molecular mechanism by which ornithine lipid synthase (OlsB) couples the non-proteinogenic amino acid ornithine to a fatty acid chain we are in the process of determining the three dimensional structure of the protein which should shed light on how it does this. To complement the analysis of the static 3D structure, modelling and molecular dynamic simulations will be performed to clarify how the enzyme functions. A second project aims to understand the mode of action of novel lanthipeptide found in a marine bacterium. Lanthipeptides are well known antibiotics and are often used in food preservation, but can also have other bioactivities. They normally exert their bactericidal activity by self assembling into pores in bacterial cell membrane leading to leakage of the cytoplasm and cell death. Molecular dynamic simulation of the new lanthipeptide in a cell membrane may give clues as to how the newly discovered peptides are similar or different to those that have been previously characterized. Currently IMBM does not have the resources to simulate these 100000+ atom systems for the length of time it requires to see meaningful changes in the protein itself or interaction with substrates. The CHPC's Lengau cluster is the fastest computer on the African continent, which supports GPU-based computation. The software (NAMD) used to model and simulate these dynamics was expressly written to take advantage of the massively parallel computing capabilities offered by GPU-compute and is one of the software packages offered by CHPC to researchers. We therefore now have access to a system which allows us, through simulation, to see these molecular interactions that may lead to robust designer enzymes for industry or altered enzymes for the production of bespoke small molecules.
Principal Investigator: Prof Emile Rugamika CHIMUSA
Institution Name: University of Kinsasha
Active Member Count: 10
Allocation Start: 2021-04-19
Allocation End: 2021-10-16
Used Hours: 6721
Project Name: African Multi-Omics Data Science
Project Shortname: CBBI1039
Discipline Name: Bioinformatics
CBBI1039: UCT Applied Genomics (AGe) is a training bioinformatics and data science platform, running from the University of Cape Town, South Africa. This research programme is fundamentally aiming 1) to provide postgraduate students with the knowledge skills and bioinformatics tools they need to understand the models of medical population genetics and of computational molecular biology; 2) to enable graduate students at the University of Cape Town, across the country and African continent to be able to relate the models and data of statistical genetics to the constraints of inheritance and the molecular mechanisms of genetic data, including DNA sequence data, next generation sequences and to develop the skills to work with these massive data sets. 3) to solicit the assistance of CHPC, South Africa, Cape Town in training the trainers along this proposed programme to efficiently use the resources and master the portable bash servers(PBS) and the implementation of large-scale genomic project under CHPC's resources. The accumulation of experimental DNA data in Biology and new high throughput experiments are growing rapidly and give a huge number of data difficult to manage, to store and to analyse this existing and future dispersed information. Therefore, high-throughput technologies, such as next-generation sequencing, have turned molecular biology into a data-intensive discipline, requiring the field of bioinformatics to use high-performance computing resources and carry out data management and analysis tasks on large scale. Today, the use of HPC has increased our training portfolio to meet the international standard with respect to large scale genomic era in training and increasing data analysis skills of our students and other trainees across the continent. This programme has developed several genomic-based, Bioinformatics, programming courses that address the African training needs around large-scale genomic data in order to meet the international standard required in this field.
Principal Investigator: Prof Cornie van Sittert
Institution Name: North-West University
Active Member Count: 24
Allocation Start: 2021-04-19
Allocation End: 2021-10-16
Used Hours: 4743814
Project Name: Computational Chemistry within LAMM at NWU (Potchefstroom)
Project Shortname: CHEM0778
Discipline Name: Chemistry
Since 1880 the main source of energy for South Africa was coal, and at present, coal provides 77% of South Africa's primary energy needs. However, electricity comes at a very high cost, namely air pollution and the influence of air pollution on human health. In 2004 the South African government reformed the legislation about air pollution. To contribute to the management of air quality, new energy sources must be investigated. These new energy sources must be affordable, clean and renewable. Some of these alternative energy sources are fuel cells and batteries. Electricity from fuel cells and batteries is obtainable through electrochemistry, where free energy of spontaneous reactions is transformed into electricity. A potential replacement form of energy is the hybrid sulphur (HyS) cycle. The HyS cycle is a two-step water-splitting process, which could be used for the production of hydrogen. In the HyS cycle, the anode material used is platinum. However, because platinum is a rare and expensive metal, it is not an economically viable option for long-term and large-scale hydrogen production. Hence, various attempts to reduce or eliminate the platinum content while not compromising the process performance has been made. To understand the electrochemistry of the anode and make informed decisions on the type and amount of metal to be used in various platinum alloys, investigations on a fundamental level are needed. These type of investigations could be classified as computational chemistry. The resources needed for these type of investigations, namely various types of hardware and software is of cardinal importance. Although resources are available at NWU, it is not nearly enough to fully support the research covered in the Laboratory of Applied Molecular Modelling (LAMM). So if we, as researchers within the LAMM, did not have access to the CHPC resources, the progress of our research will be much slower.
Principal Investigator: Dr Lonnie van Zyl
Institution Name: University of Western Cape
Active Member Count: 2
Allocation Start: 2021-04-19
Allocation End: 2021-10-16
Used Hours: 4825
Project Name: Engineering bacterial pyruvate decarboxylase for increased thermostability
Project Shortname: CBBI1265
Discipline Name: Bioinformatics
Dr. Leonardo van Zyl at the The Institute for Microbial Biotechnology and Metagenomics (IMBM) at the University of the Western Cape is investigating which portions of an enzyme contribute to its stability at high temperatures. The enzyme being investigated is called pyruvate decarboxylase (PDC) that catalyzes the conversion of pyruvate, a central metabolite in bacterial cells, to acetaldehyde. Through the action of a second enzyme called alcohol dehydrogenase, the acetaldehyde is converted to ethanol. This makes the enzymes useful in engineering bacterial strain to enable them to produce ethanol from renewable resources such as plant waste material. Bacteria which grow at high temperatures (60C), called thermophiles, have been identified as possible candidates for ethanol production, as ethanol production at these elevated temperatures has benefits for the process. The PDC enzymes are only found in organisms that grow at moderate temperatures and thus cannot withstand the high temperatures during the fermentation process. This research aims to identify the areas of the protein susceptible to high temperatures and engineer the enzyme to cope with the higher temperature. The protein crystal structures for the least- and most thermostable versions of PDC have been solved, and we can use computer simulations of the movement of these proteins at various temperatures to identify where the structure fails and address it by incorporating chemical bonds that are not so easily broken thereby retaining its functionality. A general thermostabilizing feature has not been identified for all proteins, and there is exists a huge demand for thermostable variants of a wide range of proteins globally. Should this work be successful it may suggest a general workflow for the engineering of thermostable proteins as well as shed new light on thermostabilizing features of proteins. Dr. van Zyl used the CHPC Lengau cluster to perform 250 nanosecond simulations of these medium sized (240000 atoms) protein systems at 25C and 60C to establish how temperature affects these proteins. This work should help South Africa grow its knowledge economy, display the talents of its researchers and possibly help it be a global competitor in the biofuels market through the use of publicly funded computational resources.
Principal Investigator: Dr Eric Maluta
Institution Name: University of Venda
Active Member Count: 14
Allocation Start: 2021-04-20
Allocation End: 2021-10-27
Used Hours: 166984
Project Name: Energy Material Modeling in the Application of different Photovoltaic Technologies
Project Shortname: MATS0827
Discipline Name: Physics
The Department of Physics at the University of Venda is working on understanding the properties of different materials for application in renewable energy technologies and energy storage through the density functional theory simulations as performed under the CHPC cluster. The main aim is to reduce the cost of the material used in renewable energy technologies for the future use and mitigate climate change through energy supply. We are also building capacity of the rural students on the postgraduate studies and skill development. The CHPC is a gateway for the department of Physics of the University of Venda as an institution situated in the rural area to produce researchers and to publish research work. There is a growing number of students research work from honours to the PhD level. Several student are getting their masters and PhD degree through these initiatives.
Principal Investigator: Dr Robert Luckay
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-04-20
Allocation End: 2022-02-03
Used Hours: 80092
Project Name: Computational Studies of Metal-Ligand Coordination Complexes
Project Shortname: CHEM1409
Discipline Name: Chemistry
Prof Robert Luckay heads the Ligand Design for Metal Ion Coordination in Industrial and Medical Applications Group at Stellenbosch University and is collaborating with Dr Anton Lopis at the CHPC. Most of the work performed in the group is experimental in nature, but computational work is needed to complement the experimental work and to gain a much better understanding of the Metal-Ligand (M-L) complex systems.
Improved knowledge and the development of better ligands for coordination of metal ions have various significant applications in medicine and industry. There is much knowledge to be gained from quantum calculations - such as Density Functional Theory (DFT) performed using the Gaussian code. Vital understanding of steric and electronic effects, and reaction mechanisms in such M-L complexes can be gained from DFT calculations.
Such computational knowledge would be near impossible to obtain without HPC resources such as those obtained from the CHPC. Hence CHPC is invaluable to the work in our research group.
Principal Investigator: Prof Malik Maaza
Institution Name: University of South Africa
Active Member Count: 4
Allocation Start: 2021-04-21
Allocation End: 2021-11-05
Used Hours: 480199
Project Name: Modelling of Functional Materials at the Nanoscale
Project Shortname: MATS1306
Discipline Name: Physics
The UNESCO UNISA ITL-NRF Africa Chair in Nanosciences & Nanotechnology (U2ACN2), a trilateral partnership between the UNESCO UNISA & iThemba LABS-National Research Foundation of South Africa, has established a multidisciplinary research program in materials at the nanoscale. The multidisciplinary approach of the U2ACN2 chair cements the several fragmented Africa national efforts in nanosciences and nanotechnology and addresses urgent societal needs in the water, energy, and health sectors in Africa.
Computational research uses complex models in various ways, all of which advance materials science and engineering. These computational models can help researchers understand the outcome of an experiment, identify the most promising avenues for future experiments, and give us insight into processes that cannot be easily explored in the lab. The U2ACN2 center thanks CHPC as an outstanding High-Performance Computing center, which assists our researchers by providing access to computational resources necessary to construct, analyze, and interpret their complex data in the field of nano and materials science.
Our recent projects have been focused on finding an approach via simulation methods to use the coated metal nanoparticles as deliverer with proposed drugs to treat diseases caused by the coronavirus and Malaria insect.
The Modelling computational publication on SARS-COV2 in Nature Scientific Report 2021 has been loaded & publicized on various institutional websites, UNISA, iThemba LABS & NRF. Also, the same publications are now on the open-access platform of Springer & Nature Publishing houses. Likewise, it is intended to use the same approach to investigate if ivermectin & Artemisinin phytoactive compounds can bioconjugate with nanoscaled metallic nanoparticles. The target is to study the efficacy of such bioconjugated nanosystems against Malaria.
Principal Investigator: Dr Gebremedhn Gebreyesus Hagoss
Institution Name: University of Ghana
Active Member Count: 6
Allocation Start: 2021-04-21
Allocation End: 2021-11-25
Used Hours: 197515
Project Name: Atomistic simulations and condensed matter
Project Shortname: MATS1031
Discipline Name: Physics
The condensed matter research group at the Department of Physics, University of Ghana is focused on the study of the structural, electronic, magnetic, and optical properties of 2D and bulk materials. Computational simulations of the properties of these materials are studied using density functional theory and extended Hubbard functionals as implemented in QUANTUM ESPRESSO, GW approximation of many-body perturbation theory (MBPT), and the Bethe-Salpeter equation (BSE).
Our computational work would not be possible without the enhanced computing resources available at the CHPC in South Africa. Typically, we prepare our input files in our laptops and works stations and submit the job to the CHPC cluster. The jobs are monitored at least once a day. The results are downloaded when completed.
Our main areas of research:
1. Electronic, magnetic, structural, and optical properties of transition-metal oxides. First principle calculations were carried out to study the structural, electronic, optical, and magnetic properties of transition metal oxides. The results are compared to our experimental results.
2. Perovskite Materials for energy applications The accurate description of the structural, Electronic, and optical properties of low-temperature phase perovskite are not yet well established. In this project, we focus on the investigation of structural, electronic, dielectric, and vibrational properties of low-temperature phase perovskite materials using density functional theory with the inclusion of the onsite and inter-site Hubbard correction (DFT+U+V).
3. Two-dimensional Materials Research The 2D transition metal dichalcogenides are compounds that we are interested to study in this project, have a chemical formula MX2 (where M = Ti, Mo, Nb, W, Hf, Re, etc. and X = S, Se, and Te). In particular, we focus on monolayers, bilayers, and heterostructure of transition metal disulphides (TMS2), as photo-catalysts for hydrogen evolution reaction. We are also continuing to study the excitonic properties of 2D HfS2 monolayer via lanthanide substitutional doping using GW and BSE. 4. Ruddlesden-Popper perovskite ruthenates In this project, we focus on calculating the electron-phonon couplings of Sr4Ru3O10 using DFT with the inclusion of the onsite Hubbard correction (DFT+U) and spin-orbit coupling.
Principal Investigator: Dr Jan Buys
Institution Name: University of Cape Town
Active Member Count: 11
Allocation Start: 2021-04-22
Allocation End: 2021-10-19
Used Hours: 776210
Project Name: Deep Learning for Low-resource Natural Language Processing
Project Shortname: CSCI1335
Discipline Name: Computer Science
The University of Cape Town Natural Language Processing group is a research group under the UCT Department of Computer Science that performs research on various topics related to the automated processing of language. The research group is led by Dr. Jan Buys and currently consists of 5 Masters students and 2 PhD students. Some undergraduate and honours students are also involved with the research.
The main focus is on developing methods for text generation in low-resource settings. Recent advances in Artificial Intelligence have been driven by the development of large neural networks that can "learn" how to perform various tasks by processing very large datasets - this is referred to as deep learning. For text generation, for example, deep learning systems have been developed that can automatically write realistic-looking news articles or other short pieces of text. However, deep learning techniques perform less well in settings where only small amounts of data are available. While text generation systems can perform well in English, similar systems for other South African languages (such as Nguni languages) are much less accurate or don't exist at all.
In this research project, we are performing foundational research to determine how we can adapt deep learning models to perform well in low-resource settings where only small amounts of data are available. Deep learning models are computationally intensive, even on small datasets, so they are trained on Graphical Processing Units to speed up the computations. We have developed language models, which are the models underlying deep learning text generation systems, for multiple South African languages, investigating which type of neural network is most appropriate. Using the language models as basis we have developed novel techniques to automatically subdivide words into smaller units in order to enable modelling words in morphologically rich languages, which includes most South African languages. We have also developed systems to translate text from English into various South African languages, including both the Nguni and Sotho language groups. We are investigating techniques to improve the performance given that only limited amounts of data (examples of translations) are available in these languages.
We have also developed language processing models for a number of other low-resource problems: Predicting clinical outcomes such as hospital readmission from doctor's notes; answering complex questions from documents; question answering in an interactive environment; and rewriting the style of text passages.
Principal Investigator: Prof Zenixole Tshentu
Institution Name: Nelson Mandela Metropolitan University
Active Member Count: 7
Allocation Start: 2021-04-22
Allocation End: 2021-10-19
Used Hours: 94620
Project Name: Metal-ligand, anion-cation interactions and protein-ligand studies
Project Shortname: CHEM0864
Discipline Name: Chemistry
The results obtained using CHPC facilities have allowed us to design chemistry from an informed position based on computational calculations undertaken. Without the CHPC facilities, the design of anion coordination chemistry would be done with trial and error but now it can be done rationally with computational chemistry.
Principal Investigator: Prof Edet Archibong
Institution Name: University of Namibia
Active Member Count: 7
Allocation Start: 2021-04-22
Allocation End: 2021-10-19
Used Hours: 44124
Project Name: (1) Computational Study of Semiconductor Clusters. (2) Computational Study of Bio-active Molecules Extracted from Plants
Project Shortname: CHEM0969
Discipline Name: Chemistry
Our group, The Computational Chemistry Group at the University of Namibia, are currently involved in two important projects.
For these projects, we make use of the computational resources at the CHPC in South Africa. Specifically, we use the Lengau clusters, without which we won't be able to study the systems we are currently working on. Below are brief descriptions of our work.
Our research on semiconductor clusters aim to find ways of enabling TiO2 semiconductor clusters to absorb energy efficiently in the visible region of the electromagnetic radiation. We do this by doping TiO2 with metals such as tungsten (W). This process will reduce the band gap of TiO2 making absorption in the visible region feasible. The overall benefit would be lowering the cost of solar cells which are used to scavenge solar energy for domestic and industrial use.
Our research on biologically active compounds at present involve inhibitors of enzymes that regulate cellular pH in the gonorrheae bacteria. Inhibiting the actions of these enzymes may lead to the development of new drugs that can be used for the treatment of drug resistance gohorrheae.
Both projects described above are at the embryonic stage. However, the results obtained so far are very encouraging.
Principal Investigator: Dr Babatunde J. Abiodun
Institution Name: University of Cape Town
Active Member Count: 30
Allocation Start: 2021-04-23
Allocation End: 2022-02-24
Used Hours: 3259893
Project Name: CSAG-Atmospheric Modelling
Project Shortname: ERTH0904
Discipline Name: Earth Sciences
The Climate System Analysis Group (CSAG) Atmospheric Modelling project consists of CSAG researchers and students working on model development and applications. Our group is a unique research group within Africa (CSAG http://www.csag.uct.ac.za/). We are a mix of specialties who put the needs of developing nation users at the forefront of everything we do. As a result, we seek to apply our core research to meet the knowledge needs of responding to climate variability and change. In the project, we develop, evaluate, and apply of dynamic and statistical climate models over Africa. Climate models are powerful tools for understanding the complexity of our earth climate system. We use various form of climate model, ranging from uniform-grid Global Climate Models (GCMs), Regional Climate Models (RCMs) and adaptive-grid GCMs (VGCM) that has capability to increase its grid resolution locally over a region of interest. Given the complexity and computational demand of these models we rely on high performance computers like those at the Meraka CHPC to our models. Our research goal is to understand the dynamics of climate variability (that usually induce weather and climate extremes like droughts, extreme rainfall, heatwaves, and pollution episodes), the impacts of climate change at regional scale, and how to mitigate climate change impacts in the future.
The highlight of our recently published research include: Application of MPAS () to reproduce the characteristics of the Botswana High (an atmospheric feature known for inducing drought over Southern Africa and to study the response of the high to the El El Niño Southern Oscillation (Maoyi and Abiodun, 2021 and 2022). It also includes the analysis of the potential impacts of stratospheric aerosol injection on drought risk managements over major river basins in Africa (Abiodun et al., 2021). Abiodun, et al (2021), which was selected for Nature Climate Change's highlight publication (see https://www.nature.com/articles/s41558-022-01294-9), reveal that using stratospheric aerosols injection to mitigating climate change could result in more a predictable but also potentially higher and unmanageable drought risk over large parts of Africa. The result indicates that effective emission reductions are still the safest way to avoid more droughts in Africa and could equip the African governments and stakeholders better in debating and negotiating the deployment of stratospheric aerosols injection.
Our project is making good progress. Several simulations are currently being performed or being analysed by our researchers and students using the CHPC machines.
Principal Investigator: Prof Veikko Uahengo
Institution Name: University of Namibia
Active Member Count: 5
Allocation Start: 2021-04-22
Allocation End: 2021-10-27
Used Hours: 42093
Project Name: Solar Materials
Project Shortname: MATS1043
Discipline Name: Material Science
The group from the University of Namibia (UNAM) is working on functional materials, with much emphasis on solar materials, using locally available natural resources. Research on functional materials has been on the rise lately, as a result of increase in demand for new materials used in modern application. Functional materials researched here have ranged from photocatalytic materials used in the development solar materials for making electrodes, chemosensing materials targeting the development of field testing kits for heavy metals. Solar materials have taken center stage in providing solution towards the looming energy crisis, due to increasing urban population and rural electrification. For instance, the raw materials such as zinc oxides and copper oxides are nowadays applied in the field of solar materials for harvest photons (solar materials), upon appropriately tuning of the optical properties of these oxides, while turning them into transparent thin films. Importantly, energy has become critical in our era of fast growing societies and in-demand technology, which is all energy-dependent, for technological growth and advancement. All available resources must be harnessed to fit into application of societal relevance, for value addition and the benefits of the societies. In addition, chromogenic smart materials have become a hot topic in the area of molecular recognition (such as chemosensing), in the areas of analytical chemistry and supramolecular chemistry. In this light, CHPC has become an important facility to many researchers, including our research, in cost saving for projects, in research materials (chemicals and computing softwares). Before starting with Laboratory work, simulations can guide research to streamline potential chemicals and reagents. CHPC has contributed a great deal towards saving lab time which is used to be spent on many experimental trials. Thus far, the progress of the projects has been smoothly carried out, which could not have happened without this Facility.
Principal Investigator: Dr Njabulo Gumede
Institution Name: Walter Sisulu University
Active Member Count: 1
Allocation Start: 2021-04-30
Allocation End: 2021-10-27
Used Hours: 10431
Project Name: Computer-Aided drug design for cancer therapy
Project Shortname: CHEM1058
Discipline Name: Chemistry
This research group is working with Atomwise a start-up AI/ML company from US. We are designing compounds to attack the allosteric sites of the androgen receptor. Our aim is to design antagonists that will resist mutations by not binding into the ligand binding domain of androgen receptor, that is susceptible to mutations. The lead compounds are being screened using a high-throughput screening Luminescent assay on AR+, AR-, AR null, and cell lines containing AR mutants including the ARV-7 splice variant. A one short concentration experiments have revealed 42 compounds with <30% of %cell growth. We plan to perform a 10-point calibration curve IC50 determination on these compounds.
Principal Investigator: Dr Chris Barnett
Institution Name: University of Cape Town
Active Member Count: 3
Allocation Start: 2021-04-30
Allocation End: 2021-10-27
Used Hours: 123597
Project Name: Analysis of disease related polymers
Project Shortname: CBBI1063
Discipline Name: Bioinformatics
I investigate the shapes, structures, and properties of molecules by using computational modelling techniques. I'm particularly interested in complex sugars (glycans), which are interesting to study because they are involved in intricate molecular recognition processes in cells, for example, playing a role in cell signalling, adhesion, and immunity. Further understanding of these molecules and the complex systems they are involved in may lead to the generation of additional diagnostic markers and inhibitors that can be used to combat disease. The CHPC provides a platform for being able to run large-scale and highly parallel simulations which is a great help in making rapid progress towards solving research questions. The CHPC is also a test bed when carrying out computation on and developing algorithms for large datasets.
Principal Investigator: Dr Joseph Mutemi
Institution Name: University of Nairobi
Active Member Count: 4
Allocation Start: 2021-04-30
Allocation End: 2021-10-27
Used Hours: 28241
Project Name: Numerical weather and climate modeling, prediction, forecasting and change projections for Africa and sub regions
Project Shortname: ERTH1131
Discipline Name: Earth Sciences
The ERTH1131 is led by Dr. Joseph Mutemi, Senior Lecturer at the University of Nairobi, Department of Meteorology, Kenya has been involved in addressing key research challenges on East Africa weather. On the global arena, Dr. Mutemi was a contributing author to IPCC's AR6 report. A media briefing on climate policy implications for Africa was also shared organized together with Africa Science Media Centre (AFRISMC). The postgraduate students were involved in studying key aspects of the climate aspects using the current generation of climate models, CMIP6. Dynamical biases in the Turkana jet and feedbacks between the land and atmosphere, where identified as major sources of biases on models. Such information provides model developers with useful information on regional aspects that should be improved in climate models.
Principal Investigator: Prof Ignacy Cukrowski
Institution Name: University of Pretoria
Active Member Count: 5
Allocation Start: 2021-05-03
Allocation End: 2021-11-05
Used Hours: 140901
Project Name: Understanding chemistry from QM study of molecular systems
Project Shortname: CHEM0897
Discipline Name: Chemistry
Please make use of the previously written press release – as mentioned above, no significant progress was made during the few-months-long reporting period.
Activities in chemistry (in science in general) can be seen as made of two distinctive approaches: 1) development of new theories and methodologies that should lead to better understanding of concepts and chemical process (reactions) and 2) practical implementation of known or new methodologies that should lead to routine and highly efficient experimental/computational operations. Our research group in the Department of Chemistry, University of Pretoria (Prof. Ignacy Cukrowski (the leader), Dr. Jurgens de Lange, several PhD, MSc and Honours students as well as a postdoctoral Fellows) is involved in fundamental studies. Our focus is primarily on understanding fundamentals governing inter-atomic and inter-fragment interactions from the electron density distribution throughout a molecule, or molecular system in general. For instance, there are many kinds of chemical bonds but still there is no an ultimate general theory of bonding. Hence, there are various approximate quantum chemical models that are being used to describe and explain just a specific kind of bonding. In our group we prefer to interpret classical chemical bonding, regardless of its kind, as an interaction governed by universal laws of physics. This allows us to expend the concept of interactions from classical 2-atom approach (a chemical bond) to poly-atomic interactions involving fragments of a molecular system. Such approach proved to be very useful in modelling reaction mechanisms to understand (on atomic and molecular fragment level) how and why new compounds are or are not formed as planned by a synthetic (in)organic chemist. Full understanding of successful and failed synthetic processes is of great importance as it must aid the development of, e.g., new drugs needed for treatment of (i) highly contagious diseases, such as tuberculosis, HIV/AIDS, or (ii) Alzheimer, a disease with reported cases growing rapidly throughout the world. It is important to realize that such theoretical work requires, due to the size of molecular systems under investigations, dedicated and expensive computational facilities, such as CHPC, with a dedicated staff to ensure that research groups can access and make use of computational centre 24/7.
Principal Investigator: Dr Pedro Monteiro
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 17
Allocation Start: 2021-05-03
Allocation End: 2021-10-30
Used Hours: 295753
Project Name: Southern Ocean Carbon - Climate Observatory SOCCO
Project Shortname: ERTH0834
Discipline Name: Earth Sciences
SOCCO is a CSIR-led and DSI co-funded programme that uses the comparative geographical advantage that South Africa has in the Southern Hemisphere to contribute to the global research challenge of understanding the role of the Southern Ocean in global and regional climate. The SOCCO research niche is to explore the role played by fine scale ocean dynamics in improving the climate sensitivity of ocean and coupled numerical models with a special focus in understanding the Southern Ocean's role in the CarbonClimate links. To address the scales of dynamics in the ocean, we primarily use the coupled NEMO-PISCES model in a hierarchy of resolutions from 200 km to 2 km - using the models as experimental platforms. The finer the resolution, the greater the number processes that contribute to ocean-atmosphere carbon exchange. This understanding contributes to reducing model biases in process models and climate projections as well as the development of a South African Earth Systems Model (ESM). This work advances research in global climate change based on the importance of feedbacks of the Southern Ocean dynamics on the carbon-climate system. Additionally, the ESM will be used to explore how the climate sensitivities of the Southern Ocean evolve over the 21st century as well as being South Africa's first submission to the 6th Assessment Report of the IPCC. ESM will be the basis for the medium to long term climate projections in Africa which will then support societal needs in health, water resource planning, agriculture, biodiversity management and more broadly in economic development in the region. From regional, forced ocean models to fully-coupled earth system models comprising of interacting ocean, ice, atmosphere, land, ocean biogeochemical models, numerical modelling of earth system components require massively parallel computing resources. The size and complexity of these models are such, that for SOCCO's work, the CHPC is indispensable for the resources and service it provides.
Principal Investigator: Prof Ken Craig
Institution Name: University of Pretoria
Active Member Count: 4
Allocation Start: 2021-05-05
Allocation End: 2021-11-01
Used Hours: 1030451
Project Name: CFD Modelling of Concentrated Solar Power Applications
Project Shortname: MECH0873
Discipline Name: Computational Mechanics
Central tower solar power plants are the most promising concentrated solar with research into new receivers and heat transfer fluids and collectors (heliostats) being ongoing world-wide. Researchers at the University of Pretoria's Clean Energy Research group under the leadership of Prof Ken Craig are using the resources at the CHPC to speed up calculations of both heliostat analysis (involving wind loads and fluid-structure interaction) as well as those required for a new point-focus central receiver design being developed. For heliostats, knowledge has been gained in predicting the peak loads that these reflectors will experience during all orientations and wind conditions. Ongoing work is focused on further reducing the cost of these Computational Fluid Dynamics (CFD) hybrid Large Eddy Simulation (LES):Reynolds-Averaged Navier-Stokes models to make them feasible on smaller-scale computers. This project has been ongoing since 2013 with the involvement of CHPC resources since 2016. The project has already delivered four masters students (Dawie Marais, Joshua Wolmarans, Derwalt Erasmus and Jesse Quick). PhD student Pierre Poulain is also nearing the end of this investigation into LES modelling of the atmospheric boundary layer and determining peak loads on solar collectors. For central receivers, work focused on the use of jet impingement, especially in the transient mode through either passive or active excitation. This phenomenon of heat transfer enhancement, although conceptually simple, it very hard to model accurately, and here the resources of the CHPC are invaluable, especially as LES is required to model the transient and 3D nature of swirling jet, the focus of the work of Jesse Quick. The jet impingement work is currently being taken further in another CPHC program for jet impingement boiling is used in electronics cooling.
Principal Investigator: Prof Ozlem Tastan Bishop
Institution Name: Rhodes University
Active Member Count: 8
Allocation Start: 2021-05-06
Allocation End: 2021-11-02
Used Hours: 6049179
Project Name: Structural Bioinformatics for Drug Discovery
Project Shortname: CBBI0867
Discipline Name: Bioinformatics
Research Unit in Bioinformatics (RUBi) at Rhodes University is specialized in structural bioinformatics research. Structural bioinformatics focuses on the identification and analysis of structure, dynamics and interaction of biological macromolecules in 3D space. While doing so, it aims to make connections between sequence – structure and function of macromolecules. Information on structure is particularly important as you can derive enormous amounts of information.
RUBi's main research interest is early drug discovery. Computer aided drug design technology has greatly revolutionised the drug research and development process. Traditional computational drug discovery approach includes large scale structure based and/or ligand based in silico docking combined with molecular dynamics and binding free energy calculations. All these computationally expensive calculations require to use HPC facilities.
In the last few years, RUBi has been developing some novel protocols by merging computational chemistry, structural bioinformatics and biophysics approaches and combining with traditional methods. These combined approaches also require HPC facilities. We expect that our protocols will advance our and others research. We have already been testing the combined approaches in different cases and a number of hit compounds identified; including identification of South African natural compounds as hits to cancer, malaria, trypanosoma and HIV.
Principal Investigator: Dr Ndumiso Mhlongo
Institution Name: University of KwaZulu-Natal
Active Member Count: 14
Allocation Start: 2021-05-10
Allocation End: 2021-11-06
Used Hours: 357752
Project Name: Biomolecular Modelling Research Unit
Project Shortname: HEAL1002
Discipline Name: Health Sciences
Biomolecular Modelling Research Unit is located at the University of KwaZulu-Natal, Howard College Campus and headed by Dr Ndumiso N. Mhlongo. This research group employs CHPC resources to conduct molecular dynamics simulations to analyse the dynamical properties of macro-molecules such as folding and allosteric regulations; ligand-receptor interactions; atom-to-atom interactions; protein-protein interaction;virtual screening and molecular docking of active molecules; QSAR etc, for the design and development of inhibitors against diseases and disorders. Through the use of CHPC resources, we have designed potential inhibitors against M. tuberculosis and cancer treatment. These have made contributions to discovery new inhibitors & new knowledge which could directly translate into real health benefits of the society.
Principal Investigator: Prof Giuseppe Pellicane
Institution Name: University of KwaZulu-Natal
Active Member Count: 5
Allocation Start: 2021-05-11
Allocation End: 2021-11-07
Used Hours: 345725
Project Name: Computational Study of Structure-Property Relationships in polymer blends relevant to OPV devices
Project Shortname: MATS0887
Discipline Name: Physics
Conducting polymers are organic semiconductors which are an important photovoltaic element of organic solar cells. The group leader is Dr G. Pellicane, who is a (honorary) associate professor at UKZN and NRF C1 rated scientist. Members include Mr. S. Mamba (PhD student), Prof. Tsige (Full Professor of Polymer Science at the University of Akron, US) and Prof. Workineh (Associate Professor). We work in the fields of theory and computer simulation to study the microscopic behaviour of complex fluids, and the two methods we use most are density functional theory and molecular dynamics simulations. We published several articles in scientific journals within the framework of this project, and presented our results in international conferences/workshops, that will allow for sharing and diffusion of vital knowledge with the international scientific community. We value the skillful and resourceful staff members at CHPC (Dr Anton Lopis), and are grateful to CHPC for the generous allocation of computational resources granted to us, which is instrumental to fulfill the goals of this project.
Principal Investigator: Dr Parvesh Singh
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-05-10
Allocation End: 2021-11-18
Used Hours: 8411
Project Name: Molecular dynamics and docking studies of organic compounds
Project Shortname: CHEM0795
Discipline Name: Chemistry
I am working as an Associate Professor in the School of Chemistry at the University of KwaZulu Natal (http://scp.ukzn.ac.za/Schoolleadershipandstaff2/ParveshSingh.aspx). My research focuses on the designing and synthesis of new heterocyclic assemblies with potential applications in anticancer, antibacterial, anti-TB, and anti-diabetic research. We employ molecular docking, pharmacophore modeling and molecular dynamics simulations to predict novel chemical assemblies with potential activity against these diseases. The identified lead molecules are subsequently synthesized in our synthetic laboratory and tested in vitro for their biological activities. All the computer simulations are performed on the CHPC cluster. This supercomputing facility has been very useful to gather useful information that was almost impossible to get should this facility was not available for us to use. Running these jobs on normal computers would have taken months to finish or would die in middle. With this supercomputing facility in our hands, we not only managed to run complex calculations but obtained constructive scientific explanations for our experimental results. Consequently, we managed to publish this compiled data in reputed chemistry journals. Moreover, my post-graduate students have had the opportunity to learn different computational programs installed on CHPC that are being used in the drug design and drug discovery field.
Principal Investigator: Dr Nikiwe Mhlanga
Institution Name: Mintek
Active Member Count: 13
Allocation Start: 2021-05-11
Allocation End: 2021-11-07
Used Hours: 681560
Project Name: Hydrogen Storage Materials and Catalysis
Project Shortname: MATS0799
Discipline Name: Material Science
Advanced Materials Division (AMD) houses several research and development groups; catalysis, nanotechnology (health, water purification, and sensors), and physical metallurgy. The nanotechnology platform develops diagnostic kits for diseases such as malaria, tuberculosis. The kits are in the form of field-based quantitative lateral flows and quantitative lab-based surface-enhanced Raman spectroscopy (SERS) biosensors. Understanding the precursor materials and their interaction with the analyte is of paramount importance. The HPC platform enables simulations/calculations of these systems (large systems) to gain insight into their chemistry and biology for the fabrication of better kits. Also, HPC enables the study and establishment of new memory-shaped alloys by the physical metallurgy group. Ultimately the properties inform the experimental development of the alloys envisioned for aerospace engines. The development of diagnostic kits is one of the key factors in the eradication of illnesses such as TB and malaria. And the end product is envisioned to benefit the clinical sector. The structures normally simulated for these projects are massive and require a detailed tighter computational setting to yield informative outputs. We do not have the capacity to run such calculations on our local computer hence the need for the HPC platform which we continue to benefit from.
Principal Investigator: Prof Rasheed Adeleke
Institution Name: North-West University
Active Member Count: 12
Allocation Start: 2021-05-12
Allocation End: 2021-11-08
Used Hours: 1638
Project Name: Environmental and Agricultural Microbiology
Project Shortname: CBBI1183
Discipline Name: Environmental Sciences
Our research programme at the Unit for Environmental Sciences and Management, North-West University, Potchefstroom continues to advance the understanding of microbial communities across different environmental samples, including food, water, crops, soils and agroecosystems. Our core interest is towards soil and plant health. We are actively investigating the microbial endophytes associated with ready-to-eat vegetables with the aim of identifying novel and efficient endophytes and their roles in improving plant and human health. We seek to generate novel solutions that address critical global issues, including environmental sustainability, green energy and food security. Currently, our research projects utilize state of the art sequencing technologies to decipher the whole genome of novel microorganisms, and to understand the composition and functional repertoire of microbial community in different environments such as, food, soil, water and plant. Such data can only be analysed using large computer hardware/resources. Thus, we are very reliant on the CHPC's server to execute most of our bioinformatics analysis and/or pipeline. Our research programme currently boasts of several postgraduate students at honours, masters, PhD and postdoc level level.
Principal Investigator: Dr Francois van Heerden
Institution Name: Nuclear Energy Cooperation of SA
Active Member Count: 7
Allocation Start: 2021-05-14
Allocation End: 2021-11-10
Used Hours: 8433
Project Name: Research Reactor Fuel Burnup and Material Activation
Project Shortname: INDY0788
Discipline Name: Physics
The Radiation and Reactor Theory (RRT) group at the South African Nuclear Energy Corporation (Necsa) provides calculational support for the SAFARI-1 research reactor, and perform research in the general area of nuclear analysis. We also develop and maintain our own suite of reactor simulation software, called OSCAR. This suite is the primary operational support tool for the SAFARI-1 reactor, and is used at a number of international facilities, which range from small university based research reactors, to larger national facilities. We are also involved in a number of calculational benchmarking excersises, facilitated by the International Atomic Energy Agency (IAEA).
High Performance Computing play an important role in the data preparation step for the diffusion solver available in the OSCAR-5 code suite, and also in the direct use of transport solvers to estimate neutron flux distributions in the core. It will play an even bigger role in our transition from a conservative to Best Estimate Plus Uncertainty (BEPU) calculational approach, which requires more detailed physics integration, as well as the quantification and propagation of uncertainties.
Principal Investigator: Prof Hadley Clayton
Institution Name: University of South Africa
Active Member Count: 5
Allocation Start: 2021-05-17
Allocation End: 2021-11-13
Used Hours: 61023
Project Name: Bioorganometallic Chemistry of Transition Metals
Project Shortname: CHEM1288
Discipline Name: Chemistry
Clayton Organometallic Research Group at the University of South Africa. Computer applications will be used to investigate he bioorganometallic chemistry of platinum group metals (PGMs) and their derivatives as potential metal-based anticancer drugs. Density functional methods will be applied to investigate chemical and physical properties of new transition metal complexes synthesized. Molecular docking studies are applied to study metal complex-protein interactions.
Principal Investigator: Dr Evans Benecha
Institution Name: University of South Africa
Active Member Count: 11
Allocation Start: 2021-05-18
Allocation End: 2021-11-14
Used Hours: 412624
Project Name: Defect engineering in novel 2-D materials
Project Shortname: MATS1025
Discipline Name: Physics
The research programme titled "Defect engineering in novel 2-D materials" focuses on investigation of the properties of defects in single layered materials - usually one atom thick. The group spearheading this research is headed Dr Evans Benecha and it consists of senior researchers, postdocs and graduate students from the University of South Africa. Single atom layered materials possess interesting properties as opposed to many atom thick materials. For example, Graphene – which a single layer of carbon atoms –enables electrons to flow much faster than silicon, is the world's strongest material, harder than diamond, yet lightweight and flexible. This unique combination of superior properties makes it a credible material for new technologies in a wide range of fields. Some of the envisaged applications of Graphene as well as other single atom layered materials include making of medical implants, solar cells, stronger sports equipment, and batteries; just to name a few. However, the properties of these materials are not well understood for these applications to be realized. Dr Benecha and his group's approach towards understanding these materials involves use of large computational resources, and access to the CHPC computing facility has significantly reduced the time needed to generate and analyze data. The results from this project will not only assist in the current understanding of the role of defects in these materials, but will also help to reveal potential new functionalities.
Principal Investigator: Dr Francis Opoku
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 5
Allocation Start: 2021-05-19
Allocation End: 2021-11-15
Used Hours: 1364492
Project Name: High-Throughput theoretical design of two-dimensional materials as an ultrasensitive toxic gas sensor
Project Shortname: MATS1423
Discipline Name: Chemistry
The release of anthropogenic pollutants toxic gases into the atmosphere and environment is a global concern. The use of sensor technologies raises significant legal concerns with privacy advocates and the industrial world are not immune to these changes. Currently, the world is in the very early stages of the Fourth Industrial Revolution (4IR), where new technologies are transforming manufacturing and making factories 'smart'. The underlying features of the reaction processes, such as the structural parameters and adsorption energy of the adsorbed VOCs, toxic gas and drug residues in aqueous solution, remain uncertain and such information cannot be directly achieved from experimental work. Thus, the fundamental physical driving forces, which control the reactivity of VOCs, toxic gas species and drug residues with 2D-based van der Waals heterostructure surface are still poorly understood. To address these issues, computational simulations are indispensable to offer fundamental insights to further advance the current state of knowledge. The computational laboratories of the Department of Chemistry Department and the Centre for High Performance Computing, Cape Town are equipped with most of the advanced computational software and hardware required for this study. The accumulated knowledge will be shared with the scientific community through meetings with collaborators, publications in international peer-reviewed journals and presentations at national and international conferences and workshops. It will also offer an opportunity for training and innovation as more engineering and science graduates could be involved in the production of these sensors, while those with a business background will be engaged in the distribution and marketing.
Principal Investigator: Dr Kiprono Kiptiemoi Korir
Institution Name: Moi University, Eldoret, Kenya
Active Member Count: 7
Allocation Start: 2021-05-19
Allocation End: 2021-11-25
Used Hours: 534921
Project Name: Thermal characterization of MoS2 nanostructures: an ab initio study
Project Shortname: MATS0868
Discipline Name: Material Science
Our main interest is in the area of material exploration with a focus on materials with potential applications in energy storage, ultra-hard industry, and 2D material with novel applications. In particular, increased demand for hard materials whose performance is comparable or better than that of the diamond but with low cost has necessitated the search for material with superior hardness and related properties. Niobium carbide and nitride and has been identified as a potential candidate. Single-photon emitters in hexagonal boron nitride have attracted great attention over the last few years due to their excellent optoelectronic properties. Here, through theoretical modeling, we seek to provide direct evidence that hexagonal boron nitride can be opportunely modified by introducing impurity atoms such as aluminum or phosphorus that may work as color centers for single-photon emission. Zinc oxide nanowires have been proposed as potential photo-anode materials for PEC water splitting due to their low toxicity, simple synthesis, and easy modification routes, however, zinc oxide suffers from low PEC activity and photo-corrosion effects. Therefore, the use of zinc oxide nanowires in photo-electrochemical water splitting still awaits the development of better design and synthesis procedures to improve its PEC efficiencies to commercially viable levels.
Therefore, the process of developing the aforementioned devices still poses a major technological challenge, and an in-depth understanding is required, which may lead to the development of new principles, characterization techniques, and methods that may address some of the bottle-necks currently associated with these materials.
Indeed, our recent work on the effects of defects on mechanical properties of NbC and NbN showed that the presence of anionic vacancies tends to deteriorate the mechanical properties and ultimately the mechanical hardness due to vacancy softening that can be attributed to defect induced covalent to metallic bond transition. Consequently, stringent control of anionic defects during synthesis of NbC and NbN is critical for the realization of the desired mechanical response that can make these materials ideal for super-hard and related applications.
The predictive approaches employed in our work are anchored on the principles of quantum and classical mechanics and require huge computational effort to solve the mathematical formulas, such as Schrödinger and Newtonian equations. Thus, availability of the state of the art HPC, such as CHPC is a critical component for the implementation of this work. In addition, the excellent support staff is a prerequisite for the optimal utilization of the facility whose consistent assistance has made HPC a pleasant experience.
Principal Investigator: Prof George Amolo
Institution Name: Technical University of Kenya, Nairobi, Kenya
Active Member Count: 11
Allocation Start: 2021-05-19
Allocation End: 2022-01-27
Used Hours: 954094
Project Name: Properties of Materials for Green Energy Harnessing
Project Shortname: MATS862
Discipline Name: Material Science
MATS862 is a Kenyan group employing computational simulations to study fundamental and applied materials science with a hope of building expertise to support science and technology growth in Kenya and the Eastern Africa region.
We study materials of interest to establish the properties they display as observed by other independent methods. These studies are necessary to enable us develop innovative skills that can be applied to existing or new materials in the fields of energy, water and sanitation as well as the sustainable use of the environment. Using robust computational resources, it is possible to simulate the properties of materials and extend this capability to ideal and actual materials. We have made good progress and plan to attract more young graduate students and faculty. In the last one year we are including biological molecules and bioinformatics, hoping to move towards multitheme work with chemists and molecular biologists.
The computing resources provided by the CHPC are an excellent decision support tool and considered to be highly valuable in our research work.
Principal Investigator: Dr Richard Klein
Institution Name: University of the Witwatersrand
Active Member Count: 13
Allocation Start: 2021-05-21
Allocation End: 2021-11-17
Used Hours: 73877
Project Name: PRIME
Project Shortname: CSCI1340
Discipline Name: Computer Science
PRIME is a research lab in the School of Computer Science and Applied Mathematics at Wits University. We focus on Machine Learning and Artificial Intelligence techniques for Perception and Representation Learning. The AI models that we train are very large and require weeks of computing to learn. The primary goal of this research is to be able to learn good representations of data that is unlabelled. This means that in the future, AI can learn from large unlabelled datasets which are cheaper to create, and then fine-tune their understanding of the world on smaller, more expensive, labelled datasets.
Principal Investigator: Dr Rendani Mbuvha
Institution Name: University of the Witwatersrand
Active Member Count: 2
Allocation Start: 2021-05-21
Allocation End: 2021-11-17
Used Hours: 115672
Project Name: Bayesian Methods for Neural Networks
Project Shortname: CSCI1332
Discipline Name: Data Science
We are part of a research group that is lead by Prof Tshilidzi Marwala at the Universities of Johannesburg and Witwatersrand. The group focuses on bayesian methods in machine learning. Bayesian methods allow for a principled understanding of the uncertainty around the parameters of complex methods like neural networks. Bayesian inference frequently requires the use of computationally intensive methods such as markov chain monte Carlo - that require CHPC resources to accelerate experimental time through extensive parellization. The work has results in two publications as well as numerous works in progress.
Principal Investigator: Prof Ashwil Klein
Institution Name: University of Western Cape
Active Member Count: 2
Allocation Start: 2021-05-21
Allocation End: 2021-11-17
Used Hours: 6029
Project Name: Molecular dynamic simulation
Project Shortname: CHEM1241
Discipline Name: Bioinformatics
For the various studies, we used in silico and molecular approaches including docking and MD simulation study to achieve project objectives. The tools used from the Schrodinger suite are not freely available and formed an integral part of our data analysis, therefore access to the CHPC cluster was of paramount importance. Significant progress was made with this project and the data generated from this study resulted in a manuscript published in Scientific Report with an impact factor of 4.3.
Principal Investigator: Dr James Sifuna
Institution Name: The Catholic University of Eastern Africa, Nairobi, Kenya
Active Member Count: 6
Allocation Start: 2021-05-21
Allocation End: 2022-01-27
Used Hours: 155439
Project Name: Ab initio study on novel materials for novel functionalities.
Project Shortname: MATS1424
Discipline Name: Material Science
The Theoretical Condensed Matter Group at the Catholic University of Eastern Africa, has two leaders and a couple of students. We have common interests in material discovery. I head the group at the moment and we have had a perfect rapport with every member.
The research group has a very high affinity for new novel materials that are alternatives to fossil energy.
We Employ DFT as a tool in our calculations as implimented in SIESTA and Quantum ESPRESSO codes.
CHPC is critical in the sense that it gives us computing facilities we cant achieve within our university. We study large systems that will always need plenty of CPU hours to converge. Thanks to CHPC.
At the moment, we have achieved almost half of the set objectives and we are thankful to CHPC and its technical team.
Principal Investigator: Dr Hezekiel Kumalo
Institution Name: University of KwaZulu-Natal
Active Member Count: 16
Allocation Start: 2021-05-24
Allocation End: 2021-11-20
Used Hours: 152424
Project Name: Molecular modeling and computer aided drug design
Project Shortname: HEAL1009
Discipline Name: Health Sciences
Molecular Modelling and Computational Drug Design (University of Kwazulu Natal)
The group covers a wide range of research areas in computational and molecular modelling, with a focus on biological systems and approaches to drug design.
The main interest is in the design and investigation of biologically and therapeutically oriented targets. This is achieved through the application of computational methods to the study of chemical and biochemical reactivity problems, with particular emphasis on transition state, environmental influences on mechanisms, origins of catalysis and interpretation of kinetic isotope effects. This includes mechanistic pathways and transition states for reactions in enzymes and solutions, enzyme inhibitor design and exploration of the binding and catalytic theme of designed targets, and the application of sophisticated computational approaches to understanding protein structure and function. We are involved in projects such as understanding drug resistance mechanisms using various computational tools QM /MM MD simulations Quantitative Structure-Activity Relationship (QSAR) conformational analysis of biomolecules Bioinformatics tools applications Development projects Approaches to improve the results of binding free energy calculations Development of parameters for biomolecules: ongoing projects Software implementations: ongoing projects. All the projects we undertake in the group contribute to the training of pharmaceutical scientists with specific skills for both the pharmaceutical industry and academia. This has the potential to stimulate the local pharmaceutical industry in South Africa to produce high quality and affordable medicines for optimal and cost-effective patient care, rather than depending on multinational pharmaceutical companies. We are still a young group, so funding is an issue. The resources of the CHPC allow us to develop hypotheses that we can then test experimentally much faster and with higher throughput than if we were to develop the hypotheses we want to test through experiments alone. It has given us the opportunity to explain experimental data that we could not easily explain, such as the binding landscapes of different enzymes and the mechanisms of action of different inhibitors.
Principal Investigator: Dr Michelle Gordon
Institution Name: University of KwaZulu-Natal
Active Member Count: 3
Allocation Start: 2021-05-24
Allocation End: 2022-01-18
Used Hours: 115298
Project Name: Structural Implications of Mutations in HIV-1
Project Shortname: CBBI0928
Discipline Name: Bioinformatics
Our research group is based at the Nelson R Mandela Medical School at the University of KwaZulu-Natal and focusses on the effects of mutations that cause antiretroviral drug resistance on the structure of HIV-1 proteins. The drugs that patients take as part of their treatment against HIV-1 can stop working because the virus is always changing (mutating). We are looking at how these mutations also change the structure of the virus. To do this, we make a structural model of the affected viral protein using specialised software that uses a lot of computational power. By using the CHPC, I am able to access their high powered server, as well as the software for performing the molecular modelling. With the information gained from this analysis, we can design new drugs that can work against these mutated viruses.
The project is progressing well and 5 students have already obtained their degrees using data generated at the CHPC.
Principal Investigator: Dr George Manyali
Institution Name: Masinde Muliro University of Science and Technology, Kakamega, Kenya
Active Member Count: 4
Allocation Start: 2021-05-25
Allocation End: 2021-11-25
Used Hours: 63229
Project Name: Ab Initio study of Thermoelectric Materials
Project Shortname: MATS0712
Discipline Name: Material Science
The Computational and Theoretical Physics (CTheP) Group at the Department of Physics, Masinde Muliro University of Science and Technology, comprised of a team of researchers and students dedicated to the development and application of computational modeling to outstanding problems in materials science. In general, the group seeks a deeper understanding of quantum-mechanical descriptions of interacting electrons and nuclei in order to predict and design properties of materials and devices using density functional theory as a tool for the first-principles simulation. This kind of simulation generates volumes of data that cannot be handled by the common desktop computers. High-performance computing provides a computational environment that includes parallel processing, large memory, and storage of big data. Therefore, the Center for High-Performance Computing facility in Cape Town, South Africa, plays a vital role in providing the resources that aid the discovery of new materials and accelerate its application in the industry, an achievement that would have been in vain with desktop computers. Some specific projects for our group include the characterization of already known thermoelectric materials. The information gained from this project is used as identifiers in search for new thermoelectric materials with the tailored figure-of-merit. Such materials are used to enhance the efficiency of thermoelectric generator power systems for renewable energy. Such outcomes of our projects have an impact on the livelihood of the rural communities in Kenya, and generally, support the Agenda Four of the Kenyan economic blueprint that was recently availed by the president of the Republic Kenya.
Principal Investigator: Prof Serestina Viriri
Institution Name: University of KwaZulu-Natal
Active Member Count: 15
Allocation Start: 2021-05-25
Allocation End: 2021-12-02
Used Hours: 258197
Project Name: Deep Learning in Medical Image Analysis
Project Shortname: CSCI1229
Discipline Name: Computer Science
The CHPC is providing state-of-the-art High-Performance Computing which enables to conduct of current innovative research. Powerful parallel computation processing capabilities are required in every research, especially in this era of Big Data. In our case, we have used this platform to conduct the life-saving type of research through Medical Imaging Analysis. Some positive results achieved are Accurate Skin Lesion Segmentation, Classification of Osteoarthritis Severity from Knee X-ray Images Using CNN, and even beyond medical analysis to remote sensing, etc.
Principal Investigator: Dr Lynndle Square
Institution Name: North-West University
Active Member Count: 5
Allocation Start: 2021-05-26
Allocation End: 2021-11-22
Used Hours: 79391
Project Name: Exploring poly(2,5)benzimidazole enhanced with carbon nanotubes for space applications
Project Shortname: MATS1088
Discipline Name: Physics
In this project, we are investigating a polymer nanocomposite for space applications. The has been enhanced with carbon nanotubes. The project itself has an experimental component and uses expertise across two institutions (NWU-UWC). We will explore ion bombardment, for which we will use a second national facility. The computational component for this project explores multiple scales of modelling, including fluid dynamics, Nano-scale modelling and molecular dynamics. The molecular dynamics will investigate the mechanical properties, which will be accompanied by experimental work and a second project will explore the surface ion interaction using LAMMPS. This project relies on the Material Studio license provided by CHPC and the without the computing power it provides the project would have to reconsider its approach.
Principal Investigator: Prof Ozlem Tastan Bishop
Institution Name: Rhodes University
Active Member Count: 1
Allocation Start: 2021-05-26
Allocation End: 2021-11-22
Used Hours: 4938514
Project Name: Structural Bioinformatics for Drug Discovery (2)
Project Shortname: CBBI1425
Discipline Name: Bioinformatics
Research Unit in Bioinformatics (RUBi) at Rhodes University is specialized in structural bioinformatics research. Structural bioinformatics focuses on the identification and analysis of structure, dynamics and interaction of biological macromolecules in 3D space. While doing so, it aims to make connections between sequence – structure and function of macromolecules. Information on structure is particularly important as you can derive enormous amounts of information.
RUBi's main research interest is early drug discovery. Computer aided drug design technology has greatly revolutionised the drug research and development process. Traditional computational drug discovery approach includes large scale structure based and/or ligand based in silico docking combined with molecular dynamics and binding free energy calculations. All these computationally expensive calculations require to use HPC facilities.
In the last few years, RUBi has been developing some novel protocols by merging computational chemistry, structural bioinformatics and biophysics approaches and combining with traditional methods. These combined approaches also require HPC facilities. We expect that our protocols will advance our and others research. We have already been testing the combined approaches in different cases and a number of hit compounds identified; including identification of South African natural compounds as hits to cancer, malaria, trypanosoma and HIV.
Principal Investigator: Dr Gurthwin Bosman
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-05-27
Allocation End: 2021-11-23
Used Hours: 62767
Project Name: Quantum computations for photoinduced reactions
Project Shortname: CHEM1126
Discipline Name: Physics
The research group is one of the arms of the Laser Research Institute in the Department of Physics of Stellenbosch University. Laser Research Institute of the Stellenbosch University, boasts of ongoing formidable research projects in the fields of ultrafast science, spectroscopy and laser diagnostics, laser development, trapped ion quantum control, and biophotonics and imaging, with a number of laboratories housing ultra-modern equipment. Our research group is focused on ultrafast transient absorption spectroscopy on chemical compounds ranging from simple molecules to polymers. Experimental setups for ultrafast transient absorption spectroscopy have been thoroughly developed over the years by different masters, doctoral and postdoctoral students. The research projects conducted in the ultrafast laboratory had produced masters' and doctoral theses, as well as peer-reviewed publications.
Scientific observations being made by experiments in our research group need to be properly understood and interpreted for proper communication to the general community. Therefore, a level of theoretical simulation which will allow deeper insights is needed. Part of this theoretical simulation is the density functional calculation, a theory well accepted to interpret molecular and/or structural dynamics we observe experimentally. The calculations can take forever if run on a single computer. This is where the Centre for High-Performance Computing (CHPC) comes in.
Centre for High-Performance Computing (CHPC) offers a wide range of theoretical coding and simulation with the advantage of supercomputing on parallel computers. This enables users' jobs submission to run on a faster scale. Our research group is privileged to have access to these supercomputing capabilities and the provision of simulation software for our calculations.
We are running series of density functional calculations ranging from simple molecules to polymers using Gaussian 09 that is provided by CHPC.
Principal Investigator: Prof Abu Yaya
Institution Name: University of Ghana
Active Member Count: 5
Allocation Start: 2021-05-27
Allocation End: 2022-07-14
Used Hours: 205639
Project Name: DFT modeling of weak interactions; a case study for Carbon nanosystems
Project Shortname: MATS0990
Discipline Name: Material Science
When significant amounts of carbon monoxide (CO) and carbon dioxide (CO2) gases are inhaled by humans, they can be extremely dangerous. If these gases build up in living tissues, they can lead to serious health problems such as heart disease and respiratory problems. As a result, identifying and regulating these gases when they are released in any manner is crucial. These gases can also be emitted into the world around us (schools, homes, and so on). As such, we can capture the behaviour of these gases using the Lengau computers in South Africa, by building a gas sensor that can detect their existence and prevent them from harming human tissues, which can result in illness or death.
The method used is based on the Density Functional Theory (DFT) using the Quantum Espresso code to capture atomic-scale information on the material.
Professor Abu Yaya is the principal investigator of the group of seven researchers including students from the University of Ghana, Legon. We are working on the capture of toxic gases using the CHPC in South Africa with the view of manufacturing sensors.
Principal Investigator: Dr Sharon Moeno
Institution Name: University of the Witwatersrand
Active Member Count: 2
Allocation Start: 2021-05-31
Allocation End: 2021-11-27
Used Hours: 2021
Project Name: Drug Design, Delivery and Bioactivity of Small Phenolic Compounds
Project Shortname: CHEM1287
Discipline Name: Chemistry
My group is comprised of members of the Materials unit within the Department of Oral Biological Sciences (part of the School of Oral Health Sciences) in the Faculty of Health Sciences at Wits University. Our group engages in the design of small compounds derived from phenolic compounds. The design of compounds undertakes computational analysis and in silico screening as a means of determining whether compounds possess adequate drug likeness. As part of our research activity in the unit, promising compounds are synthesized, characterized and elucidated. Finally, the bioactivity of prepared compounds is ascertained by way of conducting antimicrobial and cytotoxicity studies. This research work is conducted in an effort to identify lead compounds for possible use as antimicrobial, anticancer, and wound healing agents. The use of the CHPC clusters facilitates the in silico and computational analyses conducted by our group on a regular basis. Docking and modelling studies enabled by the use of the Lengau cluster of the CHPC enables our group to understand the interaction of the designed and prepared compounds with the identified target proteins.
Principal Investigator: Prof Rajshekhar Karpoormath
Institution Name: University of KwaZulu-Natal
Active Member Count: 14
Allocation Start: 2021-05-28
Allocation End: 2021-11-24
Used Hours: 108774
Project Name: Design of small novel organic compounds as potential drug candidates
Project Shortname: HEAL0835
Discipline Name: Health Sciences
Prof Rajshekhar Karpoormath completed his B.Pharm and M. Pharm from Karnataka University and Rajiv Gandhi University of Health Sciences, India respectively. He then obtained his PhD in Organic Chemistry (2013) from the University of KwaZulu-Natal (UKZN). In 2010 he joined UKZN as a lecturer and is currently Associate Professor in Pharmaceutical Chemistry. He is also Academic Leader of Research in College of Health Sciences, UKZN and a National Research Foundation-South Africa (NRF-SA) - Rated researcher. Prof Karpoormath started his independent research group (Synthetic and Medicinal Chemistry Research Group) in 2014. Over the years he has successfully secured several national and institutional grants worth over 6.5 million Rand (500,000 USD). These funds were utilized in establishing a well-equipped Drug Discovery Laboratory at UKZN and productive research collaboration with national and international universities as well as research institutes. His main research focus areas are Target-based drug design and synthesis of small molecules as potential anticancer and antimicrobial agents, Methodology development and development of electroanalytical methods to trace biological, organic species by fabrication/modification of electrodes with nanomaterials. His research group has published more than 100 peer-reviewed articles in international journals, two book chapters and two patents, just in a span of 7 years (2014-2021). He has successfully graduated with 21 postgraduate students (10 Ph.Ds. and 11 Masters) since 2014. He is one of the top researchers at UKZN and was the recipient of several awards in recognition of his achievements.
Principal Investigator: Dr Phillip Nyawere
Institution Name: Kabarak University, Nakuru, Kenya
Active Member Count: 6
Allocation Start: 2021-05-31
Allocation End: 2022-01-27
Used Hours: 56532
Project Name: Barium Floride and Perovskites
Project Shortname: MATS0996
Discipline Name: Physics
Barium Floride and Perovskites is a research group based in Kenya with students from four different universities. In this team are six members and I/Dr. Phillip Otieno Nyawere am the team leader.
The perovskite materials studied in this group have their applications in solar cells fabrications and battery manufacture. We have studied doped Barium Fluoride and completed this work with Elicah Wabululu. Jared is working on a superconducting material with possible high transition temperature. Such materials are proposed for conductors use of zero energy lose and will become the future conductors.
Simulations done here use quantum espresso code which is installed in CHPC. These materials use huge number of atoms or supercells which may not run completely in a laptop or desktop. The super computers are a necessity and we are indebted to appreciate CHPC for the resources provided.
One Elicah Wabululu completed her PhD last year and has been included in supervision list. Jared is completing his research work now and is working on his thesis for PhD. These two have done all their calculations in the CHPC and understand the system so well such that they are currently assigned MSc students to assist. Truphena Kipkwarkwar is graduating this December hoping to begin her PhD immediately. She has also submitted a paper to Hundawi Journal which has been accepted for publication. Perpetua has completed her simulations and she has presented in two conferences this year. Kanguha Sylvia is progressing with her simulations well.
This resource is critical for this team of young scientists because remotely they are doing their calculations that is leading to their attainments of degrees in their respective institutions of learning. This team is building up synergy in research and I believe it is becoming a team noticeable and productive in our country.
Principal Investigator: Prof Tahir Pillay
Institution Name: University of Pretoria
Active Member Count: 3
Allocation Start: 2021-06-01
Allocation End: 2021-11-28
Used Hours: 81930
Project Name: Molecular Modelling of ligands and receptors
Project Shortname: HEAL0876
Discipline Name: Health Sciences
Prof Tahir Pillay's research group in the Department of Chemical Pathology, University of Pretoria is actively working on the field of pharmcoinformatics applications in the genetic and acquired diseases. We have been using the facility of CHPC since 2014. The research group uses several pharmacoinformatics tools to design computational models followed by screening of small molecular databases and simulations for therapeutic application in several diseases including HIV/AIDS, Tuberculosis, Alzheimer's, Malaria, etc. For this purpose we are extensively using Schrodinger, Amber, Gromacs etc. Therefore, the research work is entirely dependent on pharmacoinformatics tools available in the CHPC server. The department does not have sufficient infrastructure to carry out the research in the current project. Hence, our research group requires access to the CHPC server to fulfil the objectives of the current project. We are quite happy with the services available in the CHPC server and would like to thank entire team for their efforts and support.
Principal Investigator: Prof Mattia Vaccari
Institution Name: University of Cape Town
Active Member Count: 7
Allocation Start: 2021-06-01
Allocation End: 2021-11-28
Used Hours: 19963
Project Name: HIPPO
Project Shortname: ASTR1169
Discipline Name: Astrophysics
The classification of galaxy morphology plays a crucial role in understanding galaxy formation and evolution. Traditionally, this process is done manually. The emergence of deep learning techniques has given room for the automation of this process. As such, this paper offers a comparison of deep learning architectures to determine which is best suited for optical galaxy morphology classification. Adapting the model training method proposed by Walmsley et al in 2021, the Zoobot Python library is used to train models to predict Galaxy Zoo DECaLS decision tree responses, made by volunteers, using EfficientNet B0, DenseNet121 and ResNet50 as core model architectures. The predicted results are then used to generate accuracy metrics per decision tree question to determine architecture performance. DenseNet121 was found to produce the best results, in terms of accuracy, with a reasonable training time. In future, further testing with more deep learning architectures could prove beneficial.
Principal Investigator: Dr Benjamin Victor Odari Ombwayo
Institution Name: Masinde Muliro University of Science and Technology, Kakamega, Kenya
Active Member Count: 4
Allocation Start: 2021-06-02
Allocation End: 2021-12-08
Used Hours: 18558
Project Name: Ab initio study of Solar Energy Materials
Project Shortname: MATS1426
Discipline Name: Material Science
Solar energy materials are widely researched experimentally where the materials are fabricated and characterized in a physical laboratory set up. A team of experimental and computational scientists at the Department of Physics of Masinde Muliro University of Science and Technology in Kakamega, Kenya, led by Dr Victor Odari, have purposed to incorporate computational studies of these materials to complement the experimental results. This will give a better understanding of the science and applications of the materials through the project: Ab initio study of Solar Energy Materials (MATS1426). The project aims at determining the optimum concentrations of the various elements in materials that absorb solar radiation and convert to electricity (solar absorbers) and those used in charge transport from the absorbers to the contacts of a solar cell. It also aims at determining their optimum properties such as structural, optical, defects and electrical properties by varying their doping concentrations and introduction of new elements to the materials. These results will be cost and time saving during fabrication of the materials, analyzing of results obtained experimentally and/or computationally and determining of the maximum efficiencies of various absorbers. This work is currently being done using the Quantum Espresso code in CHPC and ia also revitalizing computational physics in the department. Overall progress of the work is good and we look forward to positive results within the next allocation.
Principal Investigator: Dr Stefan du Plessis
Institution Name: Stellenbosch University
Active Member Count: 17
Allocation Start: 2021-06-03
Allocation End: 2022-03-15
Used Hours: 75407
Project Name: Shared Roots
Project Shortname: HEAL0793
Discipline Name: Health Sciences
Understanding the SHARED ROOTS of Neuropsychiatric Disorders and Modifiable Risk Factors for Cardiovascular Disease.
Shared Roots is a collaborative research project undertaken by researchers from the departments of psychiatry, neurology and genetics at Stellenbosch University and researchers from the South African National Bioinformatics Institute (SANBI). It is a MRC Flagship funded study and the principal investigator on the study is the executive head of the department of psychiatry Professor Soraya Seedat. Dr Stéfan du Plessis is responsible for the neuroimaging component. Our study included participants with posttraumatic stress disorder (PTSD), schizophrenia or Parkinson's disease. It is known that individuals with mental illness as well as other common brain disorders have higher rates of heart disease and stroke than the general population. It is not clear why this is the case. Our study therefore aims to collect information about all the potential factors that can play a role, to better understand the increased risk and enable the development of treatment strategies. To achieve this goal we will combine state-of-the-art research techniques such as genetic evaluations and the latest neuroimaging techniques. This will be combined in analysis with information evaluating the brain disorders, general health and lifestyle factors.
For our neuroimaging component, we scanned over 600 participants using Magnetic Resonance Imaging (MRI). Here we investigate potential brain changes associated with increased metabolic risk. We proceeded to calculate precise measurements of around 70 known brain regions using the MRI scans, which will be compared between cases and healthy controls. Processing takes around 24 hours for each scan. To finish in a timely manner, we used the Centre for High Performance Computing Rosebank, Cape Town, Sun Intel Lengau cluster.
Principal Investigator: Prof Willem van Otterlo
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-06-05
Allocation End: 2022-01-31
Used Hours: 1111
Project Name: Small molecule bioactives
Project Shortname: CHEM1326
Discipline Name: Chemistry
Our research group focuses on the design and synthesis of small bioactive molecules within the Group of Organic and Medicinal Organic Chemistry (GOMOC) in the Department of Chemistry and Polymer Science at Stellenbosch University. Our group utilizes the Schrodinger modelling package through the CHPC to design small molecule inhibitors of key proteins implicated in important disease states including cancer, rheumatoid arthritis and viral infections (like that from SARS-COV-2). All of these diseases have an important impact on the quality and overall productivity of human life and it would therefore be valuable to have access to other small molecules as inspiration for novel therapeutics. The access to the state-of-the-art modelling Schrodinger package is critical to the design of small libraries of compounds that are then synthesized in our laboratories, followed by bioevaluation of these molecules. Subsequent structure-activity relationships, of great value for the design of next-generation libraries, can then also be determined through this software.
Principal Investigator: Dr Katherine de Villiers
Institution Name: Stellenbosch University
Active Member Count: 4
Allocation Start: 2021-06-07
Allocation End: 2022-01-27
Used Hours: 194740
Project Name: Investigations of drugable targets relevant to antimalarial action
Project Shortname: CHEM0801
Discipline Name: Chemistry
The Bioinorganic research group (Haem Team) at Stellenbosch University (Dept. Chemistry and Polymer Science) is led by Dr Katherine de Villiers. Our primary research focus is towards understanding mechanism(s) of action of clinical antimalarials so that we are better informed when designing altogether new treatments. The latter is important because of resistance by the malaria parasite. The disease burden caused by malaria is a particular challenge in Africa, and this motivates our research efforts further. We have made use of the CHPC facilities to enable us to investigate drug-target interactions that would otherwise not be possible via experiment. In particular, we have used Materials Studio to investigate the interaction of antimalarials to the surface of crystals of synthetic malaria pigment, and Schrodinger to investigate drug-enzyme interactions.
Principal Investigator: Dr Nangamso Nyangiwe
Institution Name: Tshwane University of Technology
Active Member Count: 2
Allocation Start: 2021-06-08
Allocation End: 2021-12-05
Used Hours: 27745
Project Name: Application of density functional theory in engineered nanoparticles
Project Shortname: MATS1427
Discipline Name: Material Science
I recently started a research group at TUT Physics department. My research field is on computational materials modelling, I mostly use density functional theory (DFT) and molecular dynamic (MD) to model and predict different properties of materials. Since its arrival, the density functional theory (DFT) has experienced a substantial development, establishing nowadays as the most employed tool to tackle many different facets of the electronic structure of molecular systems. The density functional theory is, in principle, an exact method. It works with functionals, which give the energy or other properties in terms of the electron density n(r). Only a small but essential contribution to the energy, the exchange-correlation (XC) energy, thus far remains unknown as a functional of the density and has to be approximated. The overall objective of our projects is to model and predict different properties of materials so that we can assist experimentalists when they are designing their experiments.
Principal Investigator: Dr Steve Peterson
Institution Name: University of Cape Town
Active Member Count: 4
Allocation Start: 2021-06-10
Allocation End: 2021-12-07
Used Hours: 50965
Project Name: Development of a Prompt Gamma Imaging Device
Project Shortname: PHYS0830
Discipline Name: Physics
The UCT Prompt Gamma Imaging (PGI) Group is a part of the Measure Research Unit based in the Department of Physics at the University of Cape Town. PGI is a new form of imaging that will be used in conjunction with proton radiation therapy cancer treatments. The protons that are used to attack the cancer cells produce gamma-rays that we are detecting and using to produce images of the dose deposited in the patient. PGI is still a technology in development and relies heavily on computing resources for both detector design and image reconstruction.
Principal Investigator: Prof Gerrit Basson
Institution Name: Stellenbosch University
Active Member Count: 6
Allocation Start: 2021-06-10
Allocation End: 2021-12-07
Used Hours: 351045
Project Name: CFD modelling of Hydraulic Structures in Rivers
Project Shortname: MECH0985
Discipline Name: Computational Mechanics
The Institute for Water & Environmental Engineering Stellenbosch University has 3 Doctoral students and 3 Post-Doctoral student who are investigating CFD modelling of hydraulic structures with complex vortices and sedimentation. Key aspects of the research are related to the improved prediction of bridge pier scour, to investigate the addition of flares on spillways, to reduce rock bed scour in plunge pools at spillways, to optimize the design of vortex settling basins, sand traps and settlers for sediment removal at hydropower and water abstraction plants. The simulations are too computationally intensive for standard computers and the CHPC assists in the research by allowing the students to more extensively explore the hydraulic behaviour of these structures.
Principal Investigator: Dr Edwin Mapasha
Institution Name: University of Pretoria
Active Member Count: 5
Allocation Start: 2021-06-14
Allocation End: 2022-02-21
Used Hours: 148760
Project Name: Studies of defects in two dimensional materials such as graphene and tin disulphide for technological applications.
Project Shortname: MATS1429
Discipline Name: Physics
Institution: University of Pretoria
Research group: Theoretical and computational solid state research group
Project title: Studies of defects in two dimensional materials such as graphene and tin disulphide for technological applications.
Semiconducting silicon has been one of the primary materials used in the microelectronic industry for the past several decades. The silicon-based technology is nearing the limits of its use since current technologies require vastly scaled down devices. Because of this, there is a search for new, novel materials mainly two-dimensional materials that can meet this demand. Some of such two dimensional material are graphane and tin disulphide. The peculiar properties of graphane and tin disulphide include high quality Crystalinity, large surface area, high charge carrier mobilities and wide energy band gap to mention few. These unique properties ignited a large interest as a potential alternative to silicon and a candidate for various new technological applications. Some of the applications include the use of graphane for micro electronic devices, hydrogen storage (fuel cells) and as a lithium-ion battery anode. The aim of this project is to use the density functional theory methods implemented in the Quantum Espresso Package to optimize the electronic performance of graphane and SnS2 in order to facilitate its viable use in microelectronic applications. To efficiently produce reliable results we heavily rely on the higher performance computers.
Principal Investigator: Dr Michael Owen
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-06-15
Allocation End: 2021-12-12
Used Hours: 254923
Project Name: Industrial heat exchanger simulation and analysis
Project Shortname: MECH1210
Discipline Name: Computational Mechanics
The Thermofluids Division at Stellenbosch University's Department of Mechanical and Mechatronic Engineering has a long history of research in industrial heat exchangers and has conducted pioneering work in the field of dry cooling for large thermal power plants in particular.
Research in this field continues with a focus on understanding and improving the performance of these water conserving cooling systems under a variety of ambient conditions (including strong winds). Work of this nature benefits greatly from the CHPC which allows us to investigate an unprecedented number of scenarios in a short amount of time using computational fluid dynamics (CFD) and parallel computing.
Dry cooling is a key technology for South Africa's energy sector since thermal energy production (coal, nuclear, combined cycle gas and solar thermal) is typically a water intensive process which South Africa can ill afford considering the arid nature of our country. Ensuring robust operation of dry cooling systems can contribute significantly to the sustainability of South Africa's energy production.
Principal Investigator: Dr Collins Obuah
Institution Name: University of Ghana
Active Member Count: 3
Allocation Start: 2021-06-17
Allocation End: 2021-12-14
Used Hours: 2089
Project Name: Bioinorganic Chemistry
Project Shortname: CHEM1351
Discipline Name: Chemistry
The computational group at the University of Ghana is researching into finding complexes which can be use to add oxygens to compounds that contain double bonds. The produces from these transformations are important feedstock for the chemical industries to produce adhesives, surfactants, detergents etc.
From our research we have have seen that fine-tuning of the metal and its ligands will give a good catalyst for these reactions.
Principal Investigator: Prof Catharine Esterhuysen
Institution Name: Stellenbosch University
Active Member Count: 10
Allocation Start: 2021-06-17
Allocation End: 2022-02-03
Used Hours: 150145
Project Name: Computational chemistry analysis of intermolecular interactions
Project Shortname: CHEM0789
Discipline Name: Chemistry
The Computational Supramolecular Chemistry group at Stellenbosch University focuses on studying the role that intermolecular interactions play in the behaviour of a range of materials. In particular, we focus on understanding the nature and origin of the sorptive properties of porous materials with respect to the compound being adsorbed. For instance, gases such as CO2 or solvents like water undergo differing intermolecular interactions with the porous framework that can be probed through calculations performed using the CHPC's computational facility. These allow us to understand the mechanism of sorption in order to design materials with improved properties, particularly with regard to the uptake of greenhouse gases such as CO2. Such materials can thus be used to reduce the impact of the industrial carbon footprint, with the further aim of then catalysing the CO2 sequestrated within the porous framework to produce useful chemical products. Since these processes happen at molecular level they cannot fully be probed using experimental techniques, whereas with the computational facilities provided by the CHPC we can, for instance, directly follow the motion of individual gas molecules within the frameworks to explain sorption behaviour, which would be impossible using any other technique. The CHPC's facilities thus allow us to obtain a thorough understanding of the sorption behaviour that we not be able to achieve otherwise.
Principal Investigator: Prof Andrei Kolesnikov
Institution Name: Tshwane University of Technology
Active Member Count: 2
Allocation Start: 2021-06-21
Allocation End: 2021-12-18
Used Hours: 78156
Project Name: Modeling and simulation of multiphase flows with application in processing industries
Project Shortname: MECH1386
Discipline Name: Chemical Engineering
Plasma technologies related modeling of multi-phase processes is carried out at Tshwane University of Technology by group of three Master students under supervision of Prof. Andrei KOLESNIKOV
The modeling work includes application of CFD code ANSYS FLUENT for simulation of heat and mass transfer processes in thermal plasma torches, two-phase plasma-particles jets, spheroidization of refractory particles for additive manufacturing technologies.
New submodels representing momentum and heat exchange between solid and gas phase under severe conditions of plasma flows (high temperature gradient, short residence time) are under development.
High fidelity simulation of two-phase plasma flows requires modeling of millions of solid particles, while existence of high temperature and velocity gradients requires fine spatial and temporal mesh resolution.
The calculations of radiation heat transfer (method of lines are required for plasma conditions) are enormously time-consuming.
All these factors makes simulation on a 8-core desktop PC virtually impossible due to lack of speed and memory restrictions.
CHPC cluster allows to satisfy memory requirements and make computation time reasonable.
Principal Investigator: Prof Thirumala Govender
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-06-21
Allocation End: 2022-08-25
Used Hours: 23389
Project Name: IN SILICO EVALUATION OF NANO DRUG DELIVERY SYSTEMS
Project Shortname: MATS0816
Discipline Name: Material Science
The Novel Drug Delivery Unit (NDDU) at the University of KwaZulu-Natal is led by Professor Thirumala Govender, a Professor of Pharmaceutics, Head of the UKZN NanoHealth Pillar and Evaluator on the Medicine Control Council of South Africa. The NDDU currently focuses on developing advanced medicine formulations to overcome antibiotic resistance and sepsis. Antibiotic resistance is considered a global crisis currently, affects the development of human society and has high-cost implications to the government in terms of finances and resources.
Our group has designed various types of novel pharmaceutical materials as well as advanced and new generation "smart" nano-drug delivery systems such as nanomicelles, nanoplexes, polymersomes etc. with superior architectural designs which have been prepared by our team and have shown superior activity against sensitive and resistant bacteria. The group philosophy is to use a multidisciplinary integrated approach that will minimize the cost of research and maximize therapeutic outcomes.
Hence the facilities provided by CHPC allows our group to integrate molecular modelling with our research that involves extensive in vitro and in vivo animal evaluation of our novel medicines and is being recognized locally and internationally for excellence.
Principal Investigator: Prof Richard Betz
Institution Name: Nelson Mandela Metropolitan University
Active Member Count: 15
Allocation Start: 2021-06-21
Allocation End: 2022-01-27
Used Hours: 18186
Project Name: Artificial Photosynthesis
Project Shortname: CHEM0872
Discipline Name: Chemistry
Climate change caused by the emission of more and more CO2 to the atmosphere poses an existential threat to the survival of humans on this planet.
Artificial photosynthesis - as a concept - offers an additional pathway to reliably and sutainably sequester this gas from the atmosphere by turning it into useful and harmless feedstock chemicals.
A future goal is to emulate the natural photosynthesis process and produce food and oxygen independently from plants.
Principal Investigator: Prof Ken Craig
Institution Name: University of Pretoria
Active Member Count: 3
Allocation Start: 2021-06-22
Allocation End: 2021-12-19
Used Hours: 297305
Project Name: CFD modelling of jet impingement with phase change
Project Shortname: MECH1296
Discipline Name: Computational Mechanics
Jet impingement heat transfer with boiling is a promising new heat transfer method for use in power electronics cooling. Most of the current research on this topic in literature is experimental, leaving a gap for computational studies to further investigate the relevant influencing parameters. Jet impingement flow is difficult to simulate because of its transient nature and multiple flow regimes involved (from laminar to turbulent). Previous research by Prof Ken Craig's group has developed techniques based on Large Eddy Simulation to evaluate the effect of round jets with and without swirl on solar receiver heat transfer surfaces. This program extends that work to having phase change as well in the form of boiling. The latent heat absorbed by boiling works in addition to the enhance convection provided by the jet. Various case studies are being developed for application in the electronics cooling industry and the approach has been validated using test cases from literature.
Principal Investigator: Prof Richard Tia
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 37
Allocation Start: 2021-06-23
Allocation End: 2022-01-27
Used Hours: 4271684
Project Name: Computational Chemistry and Materials Science
Project Shortname: CHEM1044
Discipline Name: Chemistry
We have not yet obtained results that can form the basis of a press release but with the progress made in the our work, especially the work pertaining to the conversion and utilization of carbon dioxide, it should be possible to make a press release in the near future. With increasing interest of the scientific community and government agencies on the mitigation of rising carbon dioxide levels and the attractiveness of using carbon dioxide as a feed-stock for industrial processes, this work might be very suitable for a press release as and when we get interesting results.
Our research activities fall under two broad areas, namely;
1. Molecular modeling - Exploring the mechanisms of organic, inorganic and organometallic reactions.
These studies afford important molecular-level mechanistic details of industrially-relevant chemical reactions that are difficult, if not impossible, to obtain experimentally. Molecular level understanding enable chemical reactions to be guided quickly and efficiently along desirable pathways, with the effect of producing the right product with minimal energy consumption and minimal environmental impact. A key aspect of this work is the development of homogeneous transition metal catalysts for the depolymerization of lignin, an abundant natural resource, for fuel.
2. Materials modeling
The atomic- and molecular-level mechanistic insight gained from these studies are useful for the design of novel catalysts for the conversion of CO2 and other green-house gases as well as stranded gases from the petroleum refining into useful fuels. This addresses the problem of global warming as well as providing an environmentally friendly alternative sources of fuel.
The research objectives described here are very computationally-demanding and using the CHPC resources is highly beneficial. Without that, we would not be able to carry out some of these calculations.
Principal Investigator: Prof Orde Munro
Institution Name: University of the Witwatersrand
Active Member Count: 7
Allocation Start: 2021-06-23
Allocation End: 2022-01-27
Used Hours: 111457
Project Name: Bioinorganic Chemistry
Project Shortname: CHEM1065
Discipline Name: Chemistry
BIOINORGANIC RESEARCH GROUP – WITS UNIVERSITY
The Bioinorganic Research Group at WITS University is led by Prof. Orde Munro, who is currently the SARChI Chair of Bioinorganic Chemistry. At present, the WITS group comprises three post-doctoral research fellows (Drs. Catherine Slabber, Zeynab Fakhar, and Angelique Blanckenberg), four PhD students, and two MSc students.
NATURE OF OUR WORK
The work mainly involves the design, synthesis, and testing of novel metallodrug candidates for chemotherapy. The compounds incorporate ligands that can facilitate and control the interaction of the metal ion with biomolecules such as DNA and/or proteins. Molecular simulations are used not only to design compounds with desirable biological behavior, but also to gain fundamental insights on their mechanisms of action and electronic structures. We are particularly interested in exploiting and controlling the reactivity of both endogenous and non-natural metal ions to kill drug-resistant pathogenic bacteria such as TB (Mycobacteria) and cancer cells.
WHY WE USE SIMULATIONS
Modern drug discovery uses in silico methods for a number of reasons, including screening large libraries of compounds for binding efficacy to enzyme targets. The enzyme targets are crucial proteins for the survival and replication of cancer cells, bacteria and viruses. Screening many thousands of compounds using computer programs is an efficient way to select a handful of compounds that can ostensibly bind to the enzyme target of interest because they have the correct shape and functional groups. The small library so generated may then be synthesized in the laboratory and tested against the pathogen of interest and the putative enzyme target. In this way, computational methods shorten the time required to find promising new compounds for further testing and development as the medicines of tomorrow. Finally, computer simulations offer important three dimensional details about how drug-like compounds inhibit their enzyme targets, thereby guiding optimization of the compound's structure for enhanced target affinity and, in principle, improved in vivo efficacy.
PROJECT PROGRESS
Over the last 2 years the project has produced 2 published papers in ISI-rated international journals with another 4 papers in the pipeline. Importantly, use of the CHPC has expanded to all students in the group from MSc to post-doctoral level. The PI has also commenced use of the CHPC resources after finding the time to learn how to use the system. Regarding the overall thrust of the work, the CHPC resources have supported efforts to discover novel soluble epoxide hydrolase inhibitors, delineate the mechanism of action of gold(III) quinoline-amide topoisomerase II inhibitors, and design a library of novel metallodrug candidates for SARS-CoV-2. The latter compounds are being synthesized for in vitro screening against the virus.
Principal Investigator: Dr Uljana Hesse
Institution Name: University of Western Cape
Active Member Count: 6
Allocation Start: 2021-06-23
Allocation End: 2022-02-03
Used Hours: 185913
Project Name: Medicinal Plant Genomics
Project Shortname: CBBI1133
Discipline Name: Bioinformatics
My name is Dr Uljana Hesse, I work at the Department of Biotechnology at the University of the Western Cape in Bellville, South Africa. My research program focuses on the establishment of genome analyses of endemic South African medicinal plants locally. The program encompasses 1) generation and analysis of genome and transcriptome data from diverse endemic South African medicinal plants; and 2) development of novel computational tools for efficient storage and mining of plant sequencing data. Rooibos (Aspalathus linearis) represents the pilot plant species for the establishment of laboratorial and computational protocols. Our rooibos transcriptomes have recently been published. Current research focuses on the assembly and annotation of the rooibos genome, which involves the analysis of "big data". In previous studies, we had generated 1Tb of Illumina and 0.5 Tb of MinION sequencing data using total DNA of one rooibos genotype. After investigating different analysis approaches, we now have assembled 80% of this 1.2Gbp genome into ≈17000 contigs (N50 = 143kbp, max contig length = 1.7Mbp). Efforts are now focusing on repeat analysis and the structural and functional annotation of the genome. We have also generated first assemblies of the chloroplast and mitochondrial genomes of rooibos, and initiated transcriptome-wide differential gene expression analyses to identify genes of interest. This research requires extensive CPU resources locally only available at CHPC. It provides an outstanding opportunity for hands-on training of students in large-scale biological data analysis on a high-performance computer cluster.
Principal Investigator: Prof Hasani Chauke
Institution Name: University of Limpopo
Active Member Count: 12
Allocation Start: 2021-06-23
Allocation End: 2022-03-31
Used Hours: 194783
Project Name: Computational Modelling of Minerals, Metals and Alloys
Project Shortname: MATS1047
Discipline Name: Material Science
The minerals, metal and alloy development programme (MATS1047), led by Professor Hasani Chauke (primary investigator, PI) is amongst others one of the major strategic research niche at the Materials Modelling Centre (MMC), University of Limpopo. The work employs first-principles quantum mechanical approaches and molecular dynamics based methods, which employs various academic and commercial software with different types of interfaces. These computer-based softwares are linked to local servers (MMC) and the Centre for High Performance Computing (CHPC). The programme continues to receive enormous support from the CHPC, particularly to run large scale calculations at a more reasonable time. The programme constitutes and supports about twelve (12) postgraduate students at masters and doctoral level.
Principal Investigator: Mr Reshendren Naidoo
Institution Name: University of the Witwatersrand
Active Member Count: 2
Allocation Start: 2021-06-25
Allocation End: 2022-01-27
Used Hours: 149311
Project Name: Influence of Gas Phase Kinetics on a Pulverised Coal Flame
Project Shortname: MECH1298
Discipline Name: Computational Mechanics
The combustion of coal particles is a complex process and there are several physical processes at play. These processes are in turn highly dependent on the quality of coal being combusted and hence there must be direct link between the numerical methods employed and the true heterogeneous nature of coal. One of these facet is the way combustible material is released from coal particles. The chemical mechanisms that are used to predict the flame characteristics are important to understand the ignition/stability and emissions profile of a given combustion regime. Hence it is important to study these chemical mechanisms.
Principal Investigator: Dr Ruben Cloete
Institution Name: University of Western Cape
Active Member Count: 9
Allocation Start: 2021-06-28
Allocation End: 2022-01-27
Used Hours: 1418649
Project Name: HIV-1C integrase drug resistance
Project Shortname: CBBI1154
Discipline Name: Bioinformatics
The research group of Dr Ruben Cloete is based at the South African National Bioinformatics Institute, University of the Western Cape. The work in my group is primarily focussed on molecular modelling and drug design. Here we focus on protein structure prediction, molecular docking and simulation studies of protein-drug, protein-protein systems. Our research efforts is in understanding HIV-1 drug resistance, identifying novel drugs to treat drug resistant Tuberculosis and the prioritization of novel genes possibly associated with Parkinson's disease in South African families. This work has led to the identification of new drugs to treat Tuberculosis. Furthermore, ongoing work might also contribute to the understanding of the development of Parkinson's disease and the improved treatment of HIV-1 infected individuals within South Africa. For this to become a reality requires the use of structural computational methods to understand the binding of drugs to protein structures is key. Therefore, large scale computing resources are required to run large protein systems. Currently, we received SAMRC funding to computationally investigate SARS-CoV2 coronavirus protein targets to identify drugs and purchase the compounds to test them experimentally.
Principal Investigator: Dr Leigh Johnson
Institution Name: University of Cape Town
Active Member Count: 2
Allocation Start: 2021-06-28
Allocation End: 2022-01-27
Used Hours: 46169
Project Name: MicroCOSM: Microsimulation for the Control of South African Morbidity and Mortality
Project Shortname: HEAL1049
Discipline Name: Health Sciences
The MicroCOSM project, based at the Centre for Infectious Disease Epidemiology and Research at the University of Cape Town, is a project to simulate the spread of infectious diseases, as well as the incidence of non-infectious diseases, in the South African population. By simulating the social and biological factors that contribute to disease transmission, we can better understand which sub-populations need to be targeted for special interventions. We can also evaluate the impact that various prevention and treatment programmes have had in South Africa to date, and evaluate the potential impact of new programmes. This simulation involves generating nationally representative samples of 20,000-40,000 South Africans and tracking them over their life course. Because this involves many individual-level calculations, the model requires substantial computing power, and the CHPC is therefore critical to the conduct of these simulations. We have recently used the model to assess strategies to reduce the future incidence of cervical cancer in South Africa, and to assess the impact of different HIV programmes on trends in HIV incidence.
Principal Investigator: Dr Vuyo Mavumengwana
Institution Name: Stellenbosch University
Active Member Count: 4
Allocation Start: 2021-06-29
Allocation End: 2022-01-27
Used Hours: 7273
Project Name: Microbial symbionts bioactive compounds and their virtual screening
Project Shortname: CBBI1434
Discipline Name: Bioinformatics
The Vuyo Lab is comprised of current 3 PhD students, and 4 masters (3 upgrading to PhD) and honors students. The group is affiliated to the Stellenbosch University, Tygerberg Medical Campus. Our research aim is to mine bioactive natural products from microbial symbionts found in extremophilic environments and marine organisms. The bioactive natural products are screened in-vitro, in-vivo and virtually against pathogens infectious diseases and cancers. The work contributes to the fight against infectious diseases and cancer via producing targeted therapeutic solutions. As mentioned above we perform virtual screening of natural products via molecular dynamics simulations of complex biological systems. The simulations require robust computational power which is offered by the CHPC. This enables us to obtain results quicker and accelerates our drug discovery process. Furthermore, due to the reliability and efficiency of CHPC we have performed more simulations in a short space of time and currently drafting research papers.
Principal Investigator: Dr Kevin Lobb
Institution Name: Rhodes University
Active Member Count: 22
Allocation Start: 2021-06-29
Allocation End: 2022-01-27
Used Hours: 1735735
Project Name: Computational Mechanistic Chemistry and Drug Discovery
Project Shortname: CHEM0802
Discipline Name: Chemistry
Our work is multi-faceted, dealing both with mechanism in chemistry, and with drug discovery in the search for new active agents against disease.
In terms of mechanistic chemistry we have pipelines on github that will search smiles databases for appropriate substrates - the pipeline then generates 3-dimensional models, transition states and analyses in detail the IRC. This is a monumental set of work.
In terms of drug discovery we have lots of work in progress in the search for active agents against COVID, malaria and TB.
Principal Investigator: Prof Penny Govender
Institution Name: University of Johannesburg
Active Member Count: 6
Allocation Start: 2021-06-29
Allocation End: 2022-01-27
Used Hours: 47216
Project Name: Design of MOFs-based sorbent materials for capturing carbon dioxide
Project Shortname: CHEM0797
Discipline Name: Material Science
The computational chemistry research group at the university of Johannesburg is led by Prof Penny Govender. The research group currently focuses on material science, environmental sciences, reaction mechanisms and medicinal chemistry. Currently, Covid-19 related research has taken the center stage for our medicinal chemistry and related research. The work is being performed both for predictive purposes (where experiments are modeled before being actually carried out in the lab) and for support purposes where computational results supports experimental results. All the simulations done by the group are carried out on the Lengau cluster using software such as Materials Studio, Gaussian, Schrodinger, and AMBER, all provided by the CHPC South Africa. Our materials science research focuses on water purification and renewable energy.
Principal Investigator: Dr Lynne Shannon
Institution Name: University of Cape Town
Active Member Count: 1
Allocation Start: 2021-06-29
Allocation End: 2022-06-30
Used Hours: 1154
Project Name: Ecosystem Modelling of the Southern Benguela
Project Shortname: ENVI1388
Discipline Name: Environmental Sciences
Our research group is based at the South African Research Chair in Marine Ecology and Fisheries, University of Cape Town. Our collaborators include researchers from different South African Universities and from international universities and research institutes. Among several research areas, our work explores the use of ecosystem models to support management strategy evaluation in the Southern Benguela using different suite of models including the Atlantis modelling framework. The Atlantis model will be used to further explore climate change scenarios through comparisons between two suites of models for the southern Benguela. Progress to date has been good and further work focusing on the likely effects of climate change on the southern Benguela will continue.
Dr Kelly Ortega was selected as a co-coordinator for the regional models of the Fisheries and Marine Ecosystem models Model Intercomparison Project (Fish-MIP). Kelly will use the Atlantis model of the southern Benguela in the coming months to run simulations under the new Fish-MIP protocol focused on model evaluation under fishing and climate scenarios.
Principal Investigator: Dr Nana Ama Browne Klutse
Institution Name: Ghana Space Science and Technology Institute
Active Member Count: 3
Allocation Start: 2021-06-30
Allocation End: 2022-01-27
Used Hours: 324703
Project Name: Regional Climate modeling over Africa
Project Shortname: ERTH1087
Discipline Name: Earth Sciences
The HPC of the Center for Scientific and Industrial Research of South Africa has made our climate research possible. The research of the Regional Climate Modeling over Africa group is from different institutions including the African Institute of Mathematical Sciences, Rwanda, and the WASCAL (West African Science Service Centre on Climate Change and Adapted Land Use) and led by the University of Ghana. The research is focused on climate variability and climate change in the subregions of Africa. We use climate modeling to understand the climate system and do projections of future climate. We. do this by submitting a set of experiments on the High-performance computing system of the CSIR, running simulations of climate processes, and analyzing results. The HPC is supporting researchers and graduate students at Masters and PhD levels.
We thank the South African government and all partners involved in CHPC for making this possible. We also thank the hard-working scientists, engineers, and the general staff of the CHPC for keeping up the good services.
Principal Investigator: Prof David Lokhat
Institution Name: University of KwaZulu-Natal
Active Member Count: 1
Allocation Start: 2021-07-01
Allocation End: 2022-01-27
Used Hours: 48232
Project Name: Carbon dioxide hydrogenation
Project Shortname: CHEM1384
Discipline Name: Chemical Engineering
The project is about the development of heterogenous coagulants for water treatment which needs some background investigations. The research programme is greatly enhanced through the use of HPC. The published work and the on-going ones were faster to achieve with HPC than the local machine. There are many software packages available on CHPC which we are making use of and have fast tracked the research progress.
Principal Investigator: Prof Mahmoud soliman
Institution Name: University of KwaZulu-Natal
Active Member Count: 30
Allocation Start: 2021-07-05
Allocation End: 2022-01-27
Used Hours: 1253292
Project Name: Drug Design, development and modelling
Project Shortname: HEAL0790
Discipline Name: Health Sciences
The research scope of the Molecular Bio-Computation and Drug Design Research Lab covers a wide range of computational and molecular modeling research areas with main focus on biological systems and drug design approaches. Main interest is related to design and study of biologically and therapeutically oriented targets by employing the applications of computational methods to the study of problems of chemical and biochemical reactivity, with particular focus upon the transition state, environmental effects on mechanisms, the origins of catalysis, and the interpretation of kinetic isotope effects. This includes mechanistic pathways and transition states for reactions in enzyme and solutions; design of enzyme inhibitors and exploring the binding and catalytic theme of the designed targets and adopting sophisticated computational approaches to understand protein structures and functions. We show keen interest in diseases of global burden and of high prevalence within the south African populace such as HIV/AIDS, Tuberculosis and Cancer. The computational resources provided by CHPC is employed in performing molecular dynamic calculations and the accompanying post molecular dynamic simulation analysis.
Principal Investigator: Dr Samuel Egieyeh
Institution Name: University of Western Cape
Active Member Count: 14
Allocation Start: 2021-07-05
Allocation End: 2022-01-27
Used Hours: 8611
Project Name: Computational (Cheminformatic and Bioinformatic) Drug Discovery, Design and Development for Infectious Diseases
Project Shortname: CBBI1212
Discipline Name: Health Sciences
The Computational Pharmacology and Cheminformatics Research group aim to use cheminformatics, bioinformatics and data analytics techniques to facilitate drug discovery and development, especially from natural product, for infectious diseases.
The group hope to stimulate the drive for drug discovery and development from the rich ethnobotanical heritage of Africa.
We use various publicly available software and tools to analyze bioactivity data and predict most likely to succeed drug candidates.
The group has identified a couple of lead compounds that need to be verified with confirmatory assay.
Principal Investigator: Prof Thomas Niesler
Institution Name: Stellenbosch University
Active Member Count: 13
Allocation Start: 2021-07-05
Allocation End: 2022-02-21
Used Hours: 97871
Project Name: Automatic Speech Recognition for Under-resourced Languages
Project Shortname: CSCI1059
Discipline Name: Electrical Engineering
Most languages spoken in Sub-Saharan Africa are not technologically developed. They do not have, for example, the carefully prepared collections of speech and text that are required for the development of automatic speech recognition and associated technologies. The Digital Signal Processing Laboratory, which is part of the Department of Electrical and Electronic Engineering at Stellenbosch University, is focussing its research efforts on developing methods that allow the development of automatic speech recognisers from much smaller datasets - small enough to make their compilation feasible for the languages concerned. The developed techniques are equally applicable to South African languages, which also remain largely under resourced. For example, the group has develop the world's first pentalingual speech recognition system that is able to switch back and fourth between five South African languages (isiXhosa, isiZulu, Sesotho, Setswana and English). Such code-switching, where a speaker alternates between languages in the same utterance, is common in spontaneous South African Speech and in other multilingual situations. The development if these speech recognition systems requires the extensive training an optimisation of existing and new neural network architectures. A resource like the CHPC is essential for such computationally demanding work. The programme is progressing steadily, and has gained international attention.
Principal Investigator: Prof Christopher_W Cleghorn
Institution Name: University of the Witwatersrand
Active Member Count: 5
Allocation Start: 2021-07-05
Allocation End: 2022-02-21
Used Hours: 165698
Project Name: Population Based Metaheuristics for Machine Learning
Project Shortname: CSCI1385
Discipline Name: Computer Science
The research undertaking in the research project at WITS, focused on the hybridization of machine learning and population based approaches. This area of research is fundamentally impactful as it can be applied in areas where gradient information is not readily available, which allows a broader degree of applicably than just classic machine learning. The CHPC has allowed for this important work to proceed without being unduly contained by computational resources, and hence has empowered the students to tackle computationally challenging tasks that are of Importance to the AI community as a whole.
Principal Investigator: Prof Jeanet Conradie
Institution Name: University of the Free State
Active Member Count: 8
Allocation Start: 2021-07-05
Allocation End: 2022-01-01
Used Hours: 1911847
Project Name: Computational chemistry of transition metal complexes
Project Shortname: CHEM0947
Discipline Name: Chemistry
This project is concerned with the computational chemistry of transition metal complexes, with special focus on the application of computational chemistry in determining the structure and energy of transition metal complexes, transition states and reaction-intermediates. This is reflected in the outputs of my research group and collaborators during the period of report. The outputs gave a better understanding of experimental observation and of the factors influencing the reactivity of metal complexes, in order to streamline them for use in different applications, for example in catalysis or energy sources (batteries and solar cells), nonlinear optics, medical applications (eg HIV and anti-leukemia) and antimicrobial activity (anti- bacterial and anti-fungal).
Principal Investigator: Prof Eric van Steen
Institution Name: University of Cape Town
Active Member Count: 7
Allocation Start: 2021-07-07
Allocation End: 2022-01-27
Used Hours: 816155
Project Name: Lateral interactions on surfaces
Project Shortname: CHEM0780
Discipline Name: Chemistry
The Catalysis Institute at the University of Cape Town has developed a novel process for the direct, selective oxidation of rather unreactive methane, a major constituent of natural gas, to formaldehyde. Access the CHPC facilities was crucial here as it allowed us to determine the reaction conditions at which selective oxidation of methane could be expected. Furthermore, the reaction pathway could be explored using the CHPC facilities.
This new process opens up a novel route for the conversion of natural gas and biogas into high value chemicals.
Principal Investigator: Dr Jo-Anne de la Mare
Institution Name: Rhodes University
Active Member Count: 1
Allocation Start: 2021-07-07
Allocation End: 2022-01-03
Used Hours: 81564
Project Name: Characterization of novel inhibitors for triple negative breast cancer
Project Shortname: HEAL1391
Discipline Name: Other
The Female Cancers Research at Rhodes University (FEMCR2U) group is based in the Department of Biochemistry and Microbiology at Rhodes University. The focus of our research is preclinical drug discovery for female cancers, including triple negative breast cancer and cervical cancer. This includes the screening and characterization of novel compounds from a network of local collaborators. While drug discovery efforts in South Africa, and the continent as a whole, have historically focused on infectious diseases (ID), there is a growing appreciation of the need to focus attention towards non-communicable diseases (NCD), including cancer. This is due to the rapid rise in incidence of such diseases and predictions that the burden from NCD will overtake that of ID in the next decade. As such cancer is a neglected disease in Africa necessitating a focus on anti-cancer drug discovery that justifies the use of public resources such as the CHPC. Part of the mechanistic studies we aim to carry out for hit compounds in the early stages of the drug discovery process involve molecular docking and dynamics simulations for potential drug-target interactions, making this a crucial aspect of our drug discovery endeavors. The molecular dynamics aspect would not be possible without the use of the CHPC since we do not have the available computing power at our institution. We are very excited about the opportunity to carry out these studies, which are the first to be performed in our group. In particular, we are exploring the potential of two novel hit compounds to bind a protein domain common to a number of cancer targets, which could represent a novel approach to cancer treatment. We have completed the optimization of the simulation conditions and are now running the protein-compound simulations.
Principal Investigator: Prof Titus Msagati
Institution Name: University of South Africa
Active Member Count: 4
Allocation Start: 2021-07-08
Allocation End: 2022-01-27
Used Hours: 47816
Project Name: Metabolomics, Proteomics and Biochemistry
Project Shortname: CHEM1089
Discipline Name: Chemistry
Simple and Powerful cluster in South Africa for research students and Computer analysts.
Principal Investigator: Dr Adeola Joseph Oyenubi
Institution Name: University of the Witwatersrand
Active Member Count: 1
Allocation Start: 2021-07-09
Allocation End: 2022-01-27
Used Hours: 157169
Project Name: Optimizing balance in for Impact Evaluation
Project Shortname: ECON1213
Discipline Name: Economics and Finance
My recent work deals with portfolio optimization in cases where the optimization problem is not well behaved. This is important because in applied settings objects we are working with don't necessarily conform to our assumptions in theoretical work. Implementing heuristic optimization in a simulation study require a lot of computing power, CHPC resource makes this possible
Principal Investigator: Prof Josua Meyer
Institution Name: University of Pretoria
Active Member Count: 6
Allocation Start: 2021-07-09
Allocation End: 2022-01-27
Used Hours: 865260
Project Name: Fundamentals of forced and mixed convection in heat exchangers
Project Shortname: MECH1094
Discipline Name: Computational Mechanics
The recent trend in heat transfer enhancement with heat exchanger and its demand, has brought the academics to benchmark the mathematical non-dimensional numbers and specially the uncertainty related to it. Our current research towards thermal entrance length for a single phase flow in a close conduit, has at least convinced that it is no longer ten times the diameter, but is even the function of Reynolds and Prandtl number. With our recent investigation on cooling rate using pulse air jet impingement, once again validating the second law of thermodynamics, stated that enhancement and degradation in heat transfer, possible with use of energetic pulse jet. The CHPC simulation has helped us in defining the stage of pulse jet, where the augmentation of Enhancement/ Degradation, are observed. This is basically concerned with modification in the frequency of pulsation. The transient observation (With time and iteration), of turbulence and velocity profile, has given us few physics to define this augmentation.
Principal Investigator: Prof Mwadham Kabanda
Institution Name: University of Venda
Active Member Count: 2
Allocation Start: 2021-07-12
Allocation End: 2022-01-27
Used Hours: 227546
Project Name: Reaction mechanism for atmospheric relevant molecules
Project Shortname: CHEM1161
Discipline Name: Chemistry
This research is currently consisting of Prof MM Kabanda of the University of Venda.
The research work is focused on determining the molecular properties, reaction mechanism and kinetic properties of molecules that have both biological and atmospheric applications.
The study of molecules that possesses both biological and atmospheric or environmental applications have wide applications in different sectors of your public health as well as economic and environmental protection. for this reason, there is significant importance in ensuring that the properties of such molecular species are sell studied in order to bring to the public the awareness of their applications.
The research project is largely theoretical based and therefore requires that utilisation of computational powers that those provided by the CHPC
Principal Investigator: Mr Ernest Opoku
Institution Name: Nesvard Institute of Molecular Sciences, Ghana
Active Member Count: 8
Allocation Start: 2021-07-12
Allocation End: 2022-01-27
Used Hours: 436669
Project Name: Molecular Quantum Chemistry
Project Shortname: CHEM1352
Discipline Name: Chemistry
Nesvard Institute of Molecular Sciences is an African-focused private nonprofit research and educational institute in Ghana founded in 2019 and incorporated in 2021 under the companies Act, 2019 (Act 992). The liability of its members is limited by guarantee. Our broader objective is to propagate a new paradigm in molecular science education and research to complement the African development project.
Our vision is to advance molecular sciences in Africa through free and open world class education, training, advocacy, research, and collaboration to prepare the next generation of native African molecular scientists to solve African problems.
We aim to provide fundamental requisite skills that are less common to obtain from traditional educational institutions in Africa to complement further education. We collaborate with laboratory researchers in industry, nonprofit, government, or academic laboratories across Africa and beyond.
And even more than that, our goal has been to teach, mentor and collaborate with younger, but also more experienced native African scientists, on how to set up and perform good scientific research, write good scientific articles, and give them a springboard for further education in rewarding molecular sciences disciplines in more prestigious institutions.
Through CHPC generous computational time and resources, multiple research projects are currenting ongoing. So far, we have published 2 full length research articles, 1 accepted manuscript in press, and 2 more under review.
Principal Investigator: Prof Warren du Plessis
Institution Name: University of Pretoria
Active Member Count: 3
Allocation Start: 2021-07-13
Allocation End: 2022-01-27
Used Hours: 48039
Project Name: Electronic Warfare (EW) Technologies
Project Shortname: MECH0989
Discipline Name: Electrical Engineering
During 2021, two students completed their studies on the topic of modulation identification. Both students' work was considerably enhanced by having access to the CHPC. One student completed his fourth-year engineering project on this topic, while the other completed his masters degree. Both students used machine-learning approaches to the problem, and this is where the CHPC was so important as the relevant models took considerable time to train.
Principal Investigator: Mr Muaaz Bhamjee
Institution Name: University of Johannesburg
Active Member Count: 11
Allocation Start: 2021-07-13
Allocation End: 2022-01-27
Used Hours: 6084
Project Name: Modelling of Multiphase Flow in Process Equipment
Project Shortname: MECH0837
Discipline Name: Computational Mechanics
The research group is from the Department of Mechanical Engineering Science at the University of Johannesburg. The primary area of interest in the research is the use of coupled computational fluid dynamics and computational granular dynamics models to investigate the complex multiphase interaction in process equipment.
The models are based on multiple continuum and meso-scopic based approaches which are couple with models such as the Discrete Element Method, Immersed Boundary Method and Eulerian Dense Discrete Phase Models. Both commercial and open-source software has been used in the research. Primarily the research has been focused on
hydrocyclones. However, the research is moving onto fluidised beds, spiral concentrators, highly concentrated slurry flows and gas cyclones. The models are benchmarked accuracy) against experimental measurements as well as for computational efficiency. Accurate and computationally efficient modelling of such phenomena is vital to the design of hydrocyclones which is crucial to the mining industry. The research group has also begun to use this platform to model fluid dynamics in biomedical applications. The research group has produced numerous publications and has supported the research of four post-graduate students. Research into the computational efficiency of the models is continuing with the aid of the Centre for High Performance Computing (CHPC) facilities. Due to the complex physics and fidelity required in the problems tackled by the research group large computational resources are required. The CHPC provides a computational platform that is vital to the ongoing work and success of the research group.
Principal Investigator: Dr Moses Okpeku
Institution Name: University of KwaZulu-Natal
Active Member Count: 6
Allocation Start: 2021-07-16
Allocation End: 2022-01-27
Used Hours: 63472
Project Name: Genomics and Bioinformatics Group Westville Campus
Project Shortname: CBBI1185
Discipline Name: Bioinformatics
The Genomics and Bioinformatics Group is based in the Discipline of Genetics, University of KwaZulu-Natal, Westville. The Group is working on creating Malaria vaccine using molecular genetic approach. Malaria is a serious infectious disease endemics to most part of Africa, Asia and South America, it tis a vector borne disease responsible for the death of thousands annually. Effort of the South Africa Government have resulted in the role back of the disease from 6 of the 9 provinces of South Africa and the disease is still endemic in 3 major provinces in South Africa. Current approaches for malaria treatment have evolved into drug resistance in both parasite and host. An effective way of eradicating infectious disease is vaccination, however, available synthetic vaccines are ineffective, We are working on genetic-based vaccines. The process of vaccine development that is both time and cost-effective is to develop model proteins and simulate the reaction of the protein in humans using predictive tools. These predictive tools are either too expensive for the group to purchase or not available to small groups like ours. We depend HEAVILY on the CHPC for access to these tools which are very paramount to the success of our research.
Principal Investigator: Prof Tony Ford
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-07-19
Allocation End: 2022-02-03
Used Hours: 71531
Project Name: Computational studies of the properties of molecular complexes
Project Shortname: CHEM1134
Discipline Name: Chemistry
The research group is located at the University of KwaZulu-Natal (Ford), with a joint programme operating at Mangosuthu University of Technology (Damoyi). A wide variety of intermolecular complexes is known, some of which are extremely weakly bound. The identification of the factors which contribute to strong intermolecular binding is of fundamental importance in the understanding of the natures and structures, not only of the resulting complexes, but also of the monomers which make up these adducts. Thus, an understanding of hydrogen bonding, halogen bonding, chalcogen bonding and tetrel bonding, among others, is vital to our knowledge of the intimate properties of materials. This work requires the use of the Gaussian-16 software, which is a commercially available package fully supported by CHPC. It is an internationally used program involving many thousands of lines of code. Use of the software would be impossible without access to a high performance computing facility, such as that provided by the CHPC. Without such access this research programme would have to be abandoned.
Principal Investigator: Prof Fernando Albericio
Institution Name: University of KwaZulu-Natal
Active Member Count: 3
Allocation Start: 2021-07-20
Allocation End: 2022-01-27
Used Hours: 16492
Project Name: Peptide chemistry and Organic chemistry
Project Shortname: CHEM1102
Discipline Name: Chemistry
The project aims to be acomplished in lab. an attempt to understand the challenges faced in the lab was made using computational tools. Since, calculation of thermodynamics and kinetics require faster processors, CHPC seems quite useful. The project is ongoing.
Principal Investigator: Dr Krishna Govender
Institution Name: University of Johannesburg
Active Member Count: 6
Allocation Start: 2021-07-20
Allocation End: 2022-01-27
Used Hours: 508906
Project Name: Computational approaches to design novel anticancer agents
Project Shortname: CHEM0792
Discipline Name: Chemistry
I have been using LENGAU to run benchmarks for the next generation supercomputer that the CHPC wants to obtain.
I have also been maintaining the GPU cluster and ensuring that there is no misuse of the few GPUs that we currently have.
Without CHPC resources none of the research that I currently conduct would be possible as running simulations on my local machine would take extremely long periods of time.
Principal Investigator: Prof Thomas Scriba
Institution Name: University of Cape Town
Active Member Count: 3
Allocation Start: 2021-07-21
Allocation End: 2022-01-27
Used Hours: 227115
Project Name: Tuberculosis vaccine and biomarker development
Project Shortname: HEAL1390
Discipline Name: Health Sciences
The research programme – Tuberculosis Vaccine and Biomarker Development – is one of the research programmes within SATVI (the South African Tuberculosis Vaccine Initiative), a world leading TB research entity located within the Faculty of Health Sciences at the University of Cape Town. The research entity performs cutting edge clinical and immunological research in TB pathogenesis, biomarker development and clinical vaccine development.
One of the major obstacles in tackling TB disease is the poor understanding of the various stages in which the disease progresses in the body, and of particular interest the "subclinical" disease stage in which the individual exhibits no clinical symptoms but has radiological abnormalities or microbiological evidence of active TB disease. A deeper understanding this disease stage would make it possible to identify individuals who are in fact at this stage of the disease, although clinically asymptomatic, and treat them before they advance to the active clinically symptomatic disease state. Moreover, as some those in this subclinical stage are now known to transmit the infection, their identification and treatment would limit onward transmission of the disease.
In light of this, this particular research project aims to use transcriptomic profiling to understand subclinical and the other stages of TB disease progression in the body, and ultimately develop a diagnostic tool that can be used to tease apart individuals with subclinical TB disease from healthy ones.
The analysis of such a dataset, which is terabytes of data, is infeasible on standard laptop and desktop computers owing to memory and computing power requirements. Upon obtaining the raw data, we rely on the computing power and resources provided by CHPC to the analysis.
We are currently on course in achieving the objectives of this research project. The results of the study shall be made available to the general public in published articles.
Principal Investigator: Mr Randall Paton
Institution Name: University of the Witwatersrand
Active Member Count: 9
Allocation Start: 2021-07-21
Allocation End: 2022-01-27
Used Hours: 491909
Project Name: Compressible Gas Dynamics
Project Shortname: MECH0847
Discipline Name: Other
The CHPC provides academic researcher with access to a world-class facility to use high-performance computing in their work. The research group led by Dr Randall Paton is using these techniques in more traditional contexts, such as large-scale computational fluid dynamics modelling for complex supersonic flows. In addition to this, the research group is now looking to apply modern computing techniques deriving from machine learning to enhance the accuracy and predictive capabilities of these techniques. The students graduating from this group predominantly make use of their CHPC models to complement their experimental work and thus graduate with skills desirable in the modern, global economy.
Principal Investigator: Prof Evans Adei
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 15
Allocation Start: 2021-07-22
Allocation End: 2022-01-27
Used Hours: 343769
Project Name: Materials For Energy and Fine Chemicals
Project Shortname: CHEM1046
Discipline Name: Chemistry
There has not been any new development worth reporting since our last feedback for a press release. However, it is worth emphasizing that CHPC has become an extremely important ally in our research activities. For our students who are compelled to work with realistic molecular/material systems, the increase in their productivity when they start working on the CHPC platform becomes quite apparent. The training of our students (the future African molecular/material scientists) and our modest contribution to the understanding of molecular reaction mechanisms in the development of functional materials through our publications would not have been possible without the generous CHPC support
Principal Investigator: Prof Oleg Smirnov
Institution Name: Rhodes University
Active Member Count: 9
Allocation Start: 2021-07-22
Allocation End: 2022-03-01
Used Hours: 3235
Project Name: MeerKAT Imaging
Project Shortname: ASTR0967
Discipline Name: Astrophysics
The RATT/RARG group has been using CHPC as a testbed for distributed radio astronomy data processing applications. Future telescopes such as the SKA will required massively distributed implementations for their data processing. RATT/RARG has been using tools such as Dask and Numba to develop these, and to test them at CHPC.
Principal Investigator: Dr Pritika Ramharack
Institution Name: Medical Research Council
Active Member Count: 5
Allocation Start: 2021-07-23
Allocation End: 2022-01-19
Used Hours: 221221
Project Name: Phytomedicine in Metabolic disorders
Project Shortname: HEAL1387
Discipline Name: Health Sciences
The aim of my new Molecular modeling and Bio-computation research group within the Biomedical Research and Innovation Platform (BRIP), SAMRC, is to conduct predictive biological target identification, compound physio-chemical descriptions, molecular modeling, molecular docking and molecular dynamic simulations that are required for the enhancement of current therapeutic regimens in various metabolic disease conditions. These techniques may also be implemented in identifying and optimizing vaccine developments against SARS-CoV-2 variants amidst the Covid-19 pandemic. The group is currently still being established, with students currently being recruited for the 2021/2022 academic cycles. The work will focus on the use of the Schrodinger suite for the design and characterization of newly synthesized compounds and co-crystals, as well as the use of Glide for molecular docking. The use of the Amber suite will also be utilized to simulate a theoretical experimental environment that will be programmed using specialized chemical forcefields, thus allowing for molecular interactions and free-binding energy of the complexes to be analysed. This will provide critical information on the potential structural mechanisms of action of the compounds, as well as the structural dynamics of enzymes, with particular interest on mutational modifications. To perform these studies, the use of the CHPC will be critical in accessing the Schrondinger and Amber suites and to perform large scale molecular simulations. The successful use of the CHPC in my research is documented in various studies that are evidenced in 22 internationally peer-reviewed Journal articles (https://orcid.org/0000-0001-5850-6782?lang=en). My goal is continue utilizing this platform to facilitate and expand computational chemistry capacity development within South Africa, focusing on previously disadvantaged universities. The collaborative work within the SAMRC, using CHPC, will also provide key genomic and proteomic insights on new and circulating SARS-CoV-2 variants detected in South Africa.
Principal Investigator: Prof Charalampos (Haris) Skokos
Institution Name: University of Cape Town
Active Member Count: 5
Allocation Start: 2021-07-23
Allocation End: 2022-01-19
Used Hours: 641353
Project Name: Chaotic behavior of Hamiltonian systems
Project Shortname: CSCI1007
Discipline Name: Applied and Computational Mathematics
Using modern numerical techniques of Nonlinear Dynamics and Chaos Theory we investigate in a unified mathematical way the behavior of models describing the energy transport in disordered and granular media, as well as the properties of new elastic materials like graphene and molecules like the DNA. This research is performed by members of the 'Nonlinear Dynamics and Chaos Group' in the Department of Mathematics and Applied Mathematics at the University of Cape Town. The performed investigations constitute an innovative combination of Applied Mathematics, Chaos Theory, Hamiltonian Dynamics and Nonlinear Lattice Dynamics. Their outcomes will provide answers to some fundamental questions about the effect of chaos on the energy transport in disordered and granular media, and on the behavior of graphene and DNA. Some of the questions we try to address are: Does the presence of impurities and nonlinearities enhance or suppress the propagation of energy (e.g. heat) in solids, of light in crystals, and of vibrations in granular material? How does the ratio of the different base pairs in DNA chains affect their structural stability and the temperature at which the double-stranded DNA breaks to single-stranded DNA?
Principal Investigator: Prof Lyudmila Moskaleva
Institution Name: University of the Free State
Active Member Count: 7
Allocation Start: 2021-07-26
Allocation End: 2022-01-27
Used Hours: 253975
Project Name: Understanding surface reactivity of solids using DFT simulations
Project Shortname: CHEM1294
Discipline Name: Chemistry
The group of Prof Lyudmila Moskaleva at the University of the Free State investigates surface reactivity of solids at the atomic level using first-principles quantum-chemical methods, molecular dynamics, statistical theory, microkinetic modelling and thermodynamics. Quantum chemical methods, in particular, those based on density functional theory (DFT), are becoming increasingly popular and powerful. They can be used successfully to achieve a mechanistic understanding at the microscopic level, which is needed to optimize functional materials (such as catalysts, electrocatalysts, semiconductors, optoelectronic materials) with respect to their target properties. We would like to mention two of our ongoing projects. One of them is a computational study on the properties of a hybrid 2D system which comprises in-plane graphene and hexagonal boron nitride (h-BN) components. We have investigated the influence of graphene or h-BN domain size on the atomic geometries, electronic properties and thermodynamic properties of prototypical graphene/h-BN hybrid systems by first-principles calculations. We have demonstrated that the band gap and the thermodynamic properties can be tuned by varying the graphene or h-BN domain size. The results of this study have been recently published in FlatChem (2020) and in C (2021). Another successful project relates to the chemistry of nanoporous gold (np-Au), a versatile material possessing interesting mechanical, optical, and catalytic properties. Our study is focused on the remarkable low-temperature catalytic activity of np-Au and its high selectivity toward partial oxidation reactions. Oxidation of methanol has been studied as a model that could be extended to higher alcohols. Our theoretical study provides a detailed mechanistic picture of a complex reaction energy network on the stepped Au(321) surface from methanol co-adsorbed with O2 all the way to methyl formate partial oxidation product. In collaboration with the experimental group of Prof Risse (FU Berlin) we studied possible scenarios for the deactivation of the catalyst. To this end we investigated several pathways for oxidation of formaldehyde to formate and further to CO and H2O. We were able to show that formate is responsible for deactivation and loss of selectivity. A relevant publication is currently in preparation. We are currently learning AIMD simulations with CP2K with the goal of exploring dynamically oxidation reactions on Au-based catalysts.
Principal Investigator: Prof Shahida Moosa
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-07-26
Allocation End: 2022-01-22
Used Hours: 1152
Project Name: Rare Disease Genomics
Project Shortname: HEAL1396
Discipline Name: Health Sciences
The Rare Disease Genomics in South Africa research group is based at Stellenbosch University's Faculty of Medicine and Health Sciences. We perform next generation sequencing on samples form participants with undiagnosed rare diseases. Whole exome sequencing has been shown to be an excellent diagnostic tool in international studies. Our preliminary data show that this is certainly the case for South Africa also. With the help of the CHPC, we are able to build and optimize our analysis pipelines and provide diagnoses to the undiagnosed in southern Africa. This is improving our understanding of the genomic basis of disease in our populations. This work will be used to leverage support for broader access to exome sequencing in our setting.
Principal Investigator: Prof Peter Teske
Institution Name: University of Johannesburg
Active Member Count: 3
Allocation Start: 2021-07-26
Allocation End: 2022-02-03
Used Hours: 186669
Project Name: Coevolution of mangroves and mangrove-associated crabs
Project Shortname: CBBI0979
Discipline Name: Bioinformatics
Who?
Centre for Ecological Genomics and Wildlife Conservation, University of Johannesburg
What?
Analysis of genomic, transcriptomic and metabarcoding data of marine animals. The most noteworthy publication in 2021 was a paper in Science Advances that used genomic and transcriptomic data to explain why the KwaZulu-Natal sardine run happens. This is the highest-ranking journal ever to be published at our department (impact factor: 14.1). The paper has received much media attention, including an article in Scientific American.
Why?
To improve the management of commercially exploited species, optimise conservation efforts aimed at saving endangered species from extinction, and understand evolutionary relationships in nature.
How?
Datasets generated using massive parallel sequencing technology results can only be assembled by means of supercomputers. Using CHPC, we are able to assemble a huge number of short reads into a meaningful sequence that reflects their order in the animal genome.
Progress?
CHPC and the valuable expertise it has gathered in one place has allowed our small lab to continue punching well above its weight. Despite the challenges related to the lockdown, 2021 was even more successful than 2020, with 13 papers published (compared to 11 in 2020). This would not have been possible without the CHPC.
Principal Investigator: Dr Sunita Kruger
Institution Name: University of Johannesburg
Active Member Count: 1
Allocation Start: 2021-07-26
Allocation End: 2022-01-27
Used Hours: 51464
Project Name: Environmental Heat Transfer
Project Shortname: MECH0995
Discipline Name: Computational Mechanics
This research group in the Mechanical Engineering Science Department of the University of Johannesburg is currently focusing on environmental heat transfer. Specifically heat transfer in naturally ventilated greenhouses. Greenhouses are used to protect plants from adverse weather conditions and insects. Ventilation of greenhouses are of vital importance to ensure quality crop production. If temperatures in a greenhouse is too high, poor plant growth may result, and an increased need for frequent watering. A mechanical ventilation system might be required to cool the inside of the greenhouse. Natural ventilation is an alternative option used to ventilate greenhouses. Natural ventilation uses temperature and wind to control the indoor climate of greenhouses. Unfortunately greenhouses are extremely energy intensive. Energy costs are the third highest cost related to greenhouse crop cultivation. Reducing the operating costs of energy associated with greenhouse cultivation may result in a price reduction of greenhouse cultivated crops. Conducting experimental work on ventilation of greenhouses can be costly and cumbersome. Using computational methods such as CFD (Computational Fluid Dynamics) to obtain qualitative and quantitative assessment of greenhouses can reduce costs and time involved. The computer cluster at the Centre for High Performance Computing has been used to conduct the numerical investigation using CFD. Currently research is being conducted on a large rooftop greenhouse. Smaller greenhouses containing a single and multi-spans have also been investigated. Currently, three dimensional models of greenhouses subject to buoyancy driven flow are being investigated.
Principal Investigator: Prof Rebecca Garland
Institution Name: University of Pretoria
Active Member Count: 21
Allocation Start: 2021-07-27
Allocation End: 2022-02-03
Used Hours: 667226
Project Name: Development of the first African-based earth system model VRESM and its projections of future climate change over Africa
Project Shortname: ERTH0859
Discipline Name: Earth Sciences
The first African-based Earth System Model (ESM) is under development at the Council for Scientific and Industrial Research (CSIR) in South Africa, through collaboration with the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia and South Africa's Centre for High Performance Computing (CHPC). The new ESM is unique in the sense that it is being developed in Africa and through the lens of African and Southern Hemisphere climate processes. That is, the reliable simulation of climate issues critical to Africa, such as the occurrence of drought in in response to El Niño events, wide-spread flooding over Mozambique due to landfalling tropical cyclones and increased occurrences of heat-waves under global warming are key issues in the development of the new model. The model development process is also focussed on the Southern Ocean's role in regulating southern African climate, as well as the global climate through the absorption of carbon dioxide in the ocean. The development and application of an Earth System Model is a computationally extensive endeavour, and was impossible to undertake in South Africa before the Lengau cluster of the CHPC became available in 2018. The CSIR-CHPC partnership towards development of the new model has in 2018 led to the most detailed set of projections of future climate change over southern Africa obtained to date. These simulations have provided new insight into the plausible impacts of climate change in South Africa, including how the frequencies of droughts, heat-waves and landfalling tropical cyclones may change over the next few decades. These projections were key in informing South Africa's Third National Communication on Climate Change, which was published in 2018 under the science leadership of the CSIR. Moreover, the research papers based on the projections have directly informed the Special Report on Global Warming of 1.5 ºC of the Intergovernmental Panel on Climate Change (IPCC).
Principal Investigator: Dr Ndanduleni Lethole
Institution Name: University of Fort Hare
Active Member Count: 4
Allocation Start: 2021-07-27
Allocation End: 2022-02-03
Used Hours: 306542
Project Name: Computational Studies of MxPt1-x (M; Mn, Fe, Co and Ni) alloys for magnetic data storage and biomedical applications.
Project Shortname: MATS1309
Discipline Name: Material Science
The research group is located under the department of Physics in the University of Fort Hare. The group started in February 2021 and is composed of the Principal Investigator and two BSc Honours students. Currently, the group entirely depends on the Centre of High Performance Computing (CHPC) facility since it uses Material Studio modeling and simulation environment accessed via the CHPC as its only simulation package. The group aims to grow to at least two MSc and 2 Honours students in the year 2022 and also acquire a license for the Vienna ab initio simulation package. The group is currently undertaking two computer simulation studies, namely; first-principles calculations of M-Pt (M: Mn, Fe, Co and Ni) alloys for potential application as advanced performance permanent magnets, new storage of ultra-high-density magnetic data and in biomedical applications; and Zn/CaMn2O4 for potential application in Zn and Ca ions rechargeable batteries. The M-Pt alloys is significant since there has been a remarkable increase in data storage density of magnetic disk drives in the past few decades which is projected to be restricted by the "super-paramagnetic limit" (loss of data due to activated fluctuations of magnetization) in the near future. Super-paramagnetic limit can be overcome by designing new magnetic storage materials which are energetical, electronical, magnetical, mechanical and dynamical stable. Thanks to modern simulation packages and sufficient computing resources available at the CHPC, augmentation of various properties of the crystal structure is possible by theoretically calculating the forces acting on the nuclei. Simulations are set up on the local desktop computer using the BIOVIA Materials Studio modeling and simulation environment and submitted to the CHPC computing resources using the CASTEP code.
Principal Investigator: Dr Madison Lasich
Institution Name: Mangosuthu University of Technology
Active Member Count: 10
Allocation Start: 2021-07-27
Allocation End: 2022-01-27
Used Hours: 66621
Project Name: Chem Thermo
Project Shortname: CHEM1028
Discipline Name: Chemical Engineering
The Thermodynamics, Materials and Separations Research Group at Mangosuthu University of Technology has been using the high performance computing facilities of the CHPC to investigate the behaviour of materials at the molecular level. A key aim of the research group is to enable the use of common materials for novel applications, especially to enhance energy efficiency in industrial settings. Thus far the research group has produced 8 peer-reviewed publications, including potential novel applications of waste concrete.
Principal Investigator: Prof Marelie Davel
Institution Name: North-West University
Active Member Count: 10
Allocation Start: 2021-07-26
Allocation End: 2022-01-27
Used Hours: 51056
Project Name: DNN Analysis
Project Shortname: CSCI1191
Discipline Name: Computer Science
MUST is a research niche area that performs basic and applied research in machine learning (DL), with an emphasis on the theory and application of DL. Our theoretical work focuses on generalisation in deep learning, and the interpretability of deep learning models. Building on a strong track record in the application of machine learning to multilingual speech processing, our current application domains are diverse, ranging from speech processing to space weather prediction to industrial applications of deep learning.
In the past decade, the field of Deep Neural Networks (DNNs) has brought renewed energy and focus to AI, through a series of remarkable breakthroughs in fields as diverse as speech recognition, board games and self-driving cars. In these and other applications, DNN systems have reached previously unknown levels of accuracy, making human-level performance a distinct possibility and thus suggesting novel insights on the mind-matter problem.
The successes of DNN systems have inspired much research into better algorithms, novel applications and a better understanding of DNNs. The MuST group is involved in all these aspects of DNN research. For example, we are using DNNs and word embeddings to develop better language models for under-resourced languages; these models can be used in tasks such as speech recognition and machine translation. We are also using DNNs to handle poor quality audio in speech and speaker recognition systems better, probe the processes at play during solar flare eruptions, and even optimising the design process of airfoil shape with some of our industry partners. We balance these applications with theoretical work focussed on understanding and characterising generalisation in the context of deep learning.
MuST hosts the CAIR Deep Learning group of the Centre for Artificial Intelligence Research (CAIR), an initiative aimed at developing world-class AI research leadership and teaching capability in South Africa.
Principal Investigator: Dr Nkululeko Emmanuel Damoyi
Institution Name: Mangosuthu University of Technology
Active Member Count: 2
Allocation Start: 2021-07-30
Allocation End: 2022-04-28
Used Hours: 26808
Project Name: Surface Reaction Mechanisms
Project Shortname: CHEM1105
Discipline Name: Chemistry
My present area of research is in Computational Chemistry (CC) modelling. Computer technological advances make it possible to do research in CC in order to model chemical reactions and predict chemical reaction properties which are difficult or impossible to determine from normal laboratory experiments. Most of my present research involves utilizing CC to predict the likely chemical reaction mechanisms that produce valuable organic compounds from alkanes with use of catalysts and also without catalysts.
Principal Investigator: Prof Raymond Hewer
Institution Name: University of KwaZulu-Natal
Active Member Count: 8
Allocation Start: 2021-07-29
Allocation End: 2022-01-25
Used Hours: 6704
Project Name: The design and discovery of small-molecule inhibitors of HIV-1 integrase activity
Project Shortname: CHEM0877
Discipline Name: Chemistry
Our research group, led by Dr Raymond Hewer, within the Discipline of Biochemistry at the University of KwaZulu Natal seeks to design and discover novel compounds with therapeutic potential against HIV-1 and Alzheimer's Disease; two diseases that represent a high burden of disease in South Africa. Our primary approach to identify new compounds involves the use of state-of-art molecular modelling programs which we access through agreement with the CHPC. A number of compounds identified in this manner have proven to be effective in vitro with findings published in various Journal articles
Principal Investigator: Dr Jaap Hoffmann
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-07-29
Allocation End: 2022-01-25
Used Hours: 217842
Project Name: Flow through porous media
Project Shortname: MECH1116
Discipline Name: Computational Mechanics
Unlike most other renewable electricity sources, solar thermal power plants with thermal energy storage are capable to generate electricity 24/7.
Gas turbine (Brayton) cycle is attractive, since at high temperatures and compression ratios, its thermal efficiency is higher than most other thermodynamic cycles. Packed beds find application as thermal storage for a solarized Brayton cycle, as it use cheap materials (rocks) and harmless gases (air) in its design. As a result, it is much cheaper than other storage technologies. Notwithstanding, proper design calls for a good understanding of the heat transfer and pressure drop in the rock bed to design the most effective storage systems. Our research aims to help designers come up with the best possible rock bed designs.
Principal Investigator: Prof Linky Makgahlela
Institution Name: Agricultural Research Council
Active Member Count: 8
Allocation Start: 2021-07-29
Allocation End: 2022-01-25
Used Hours: 34552
Project Name: Beef and Dairy Genomics
Project Shortname: CBBI1138
Discipline Name: Bioinformatics
The ARC-Animal Production, Animal Breeding and Genetics unit is based in Irene, Pretoria. Our research activities aim to develop knowledge on the animal genetic resources of South Africa, and apply traditional and genomics technologies for improved climate-smart production in the livestock industry while preserving animal biodiversity. The growing global population threatens food security, and half of this growth is expected in Africa. Climate change will reduce available natural resources needed for agriculture. We need to produce 100% more food, than today, in the next 30 years, and with limited natural resources. Genomic technological innovations promises to deliver genetic diversity, efficiency, product quality, food safety and public health, animal health and welfare and environment, all interlinked. Current research develops strategies for characterization, conservation and utilization of indigenous animal (cattle, sheep & goats) genetic resources. These are adapted genotypes that will survive global warming. Combine advance genomic technology data (DNA markers and whole genome sequence) and traditional animal pedigrees and longitudinal performance data, and develop statistical models for national genetic evaluations estimated breeding values. The use of genomically enhanced breeding values will speed up the rate of identifying the breeding stock and food production in cattle, sheep and goat sectors.
All research activities are carried out in collaboration with partners e.g. SA Universities, Commodity groups and Government departments. Post-graduate students are linked to the research as the bloodline for capacity building. Beef genomic research results in inbreeding and imputation (MSc. and PhD degrees awarded) have been disseminated to the cattle breeders association to facilitate uptake. Dairy genomic research is ongoing, and all projects utilizing whole genome sequences are ongoing.
Principal Investigator: Dr Saheed Sabiu
Institution Name: Durban University of Technology
Active Member Count: 8
Allocation Start: 2021-07-31
Allocation End: 2022-01-27
Used Hours: 544422
Project Name: Drug Discovery & Development and Viral Metagenomics
Project Shortname: HEAL1361
Discipline Name: Bioinformatics
The Drug Discovery & Development and Viral Metagenomics group at Durban University of Technology, looks into the concepts of healing with plant-derived compounds while reporting health benefits in a way that will provide valuable data which will lead to new drug discovery.
Besides drug discovery interests, the group is also focusing on molecular dynamics of enteric and respiratory viruses using whole genome sequencing and metagenomic approaches.
Principal Investigator: Prof Liliana Mammino
Institution Name: University of Venda
Active Member Count: 3
Allocation Start: 2021-08-02
Allocation End: 2022-01-29
Used Hours: 502259
Project Name: computational study of biologically active molecules of natural origin
Project Shortname: CHEM0959
Discipline Name: Chemistry
WHO ARE WE?
One professor, one student who has just completed his M.Sc. degree, and one student now starting her third Ph.D. year (at the University of Venda).
WHAT DO WE DO?
We focus on the study of biologically active molecule. These are molecules that can be interesting for the development of new drugs to treat diseases. We are currently studying molecules that are active against diabetes, cancer, malaria and some viruses (including one that has been reported to be active against covid). We study molecules of natural origin because their biological activity is already proven and they constitute better 'starting points' for drug development.
WHY IS IT IMPORTANT TO STUDY MOLECULES?
Studying molecules computationally means making calculations to find their properties.
All the properties of substances depend on the properties of their molecules, including pharmacological activities. We study molecules from natural sources. It may be possible that some modified molecules have stronger activity. Specialists who work on further steps of drug development can use the information from our studies to select the most suitable molecules and predict the activities of various modified molecules.
Drug development is not a fast process. It requires a lot of contributions from different specialists. Our contribution is that of providing information about the properties of promising molecules.
HOW?
We use calculations to find the properties of the molecules that we consider.
Most of these molecules are not very small, and the calculation of their properties requires huge computer power. Without the use of the CHPC, it would not be possible to obtain most of the results that we need. That is why using the CHPC is essential for us to be able to conduct our research.
HOW IS THE PROJECT PROGRESSING?
The project is progressing satisfactorily, despite some challenges caused by the pandemic.
Principal Investigator: Dr Rian Pierneef
Institution Name: University of Pretoria
Active Member Count: 15
Allocation Start: 2021-08-02
Allocation End: 2022-01-29
Used Hours: 51835
Project Name: Bioinformatic and Computational Biology analyses of organisms
Project Shortname: CBBI1124
Discipline Name: Bioinformatics
The SARS-CoV-2 pandemic has clearly illustrated the importance of genomics and next-generation sequencing in our daily lives. This invaluable tool has enabled rapid response to new variants and aided COVID-19 policy and monitoring. The essence of this project is based on the production and analysis of next-generation sequencing data across various environments and value chains, in particular the Agricultural sector. Access to the CHPC enables members of this project to efficiently handle large data sets and provides them a platform with the required analysis tools to accomplish their research objectives.
Principal Investigator: Prof Juliet Pulliam
Institution Name: Stellenbosch University
Active Member Count: 18
Allocation Start: 2021-08-02
Allocation End: 2022-01-29
Used Hours: 28631
Project Name: South African Centre for Epidemiological Modelling and Analysis
Project Shortname: CBBI1106
Discipline Name: Other
SACEMA is a DSI-NRF Centre of Excellence focused on Epidemiological Modelling and Analysis, hosted at Stellenbosch University. We use computational approaches to address questions of public health relevance to South Africa and the African continent. We use the CHPC primarily to run stochastic simulations and for simulation-based validation of novel inference methods.
Principal Investigator: Prof Yoshan Moodley
Institution Name: University of Venda
Active Member Count: 6
Allocation Start: 2021-08-03
Allocation End: 2022-01-30
Used Hours: 181629
Project Name: Mammalian Evolutionary Genomics
Project Shortname: CBBI0911
Discipline Name: Bioinformatics
This research group based at the University of Venda in the tropical far north of South Africa, focusses on reconstructing evolutionary history from whole genome sequences. There is less than a handful of other institutions in Africa that are able to carry out this line of research. There are various reasons for this. The last 10 years has seen a massive paradigm shift in biology towards big data research, spurred on by next generation (genome scale) sequencing technologies as well as other omics strategies and the availability of large data crunching servers like the CHPC. RSA has been slow in both the generation and the analysis of genome wide data, with the greatest advances coming from the medical and agricultural fields, with little development in the zoological and evolutionary contexts. This group breaks that mold. There are currently several projects being run under this theme by my postgraduate and postdoc researchers. Due to the big data nature of these projects, it is absolutely essential that users have access to an external resource like the CHPC, as this not only provides high performance computing capabilities, but also a "safe and reliable" off campus place to run our analyses, which is independent of the dysfunctionality that otherwise characterizes the University of Venda.
Principal Investigator: Prof Mario Santos
Institution Name: University of Western Cape
Active Member Count: 18
Allocation Start: 2021-08-03
Allocation End: 2022-01-30
Used Hours: 1359109
Project Name: Cosmology with Radio Telescopes
Project Shortname: ASTR0945
Discipline Name: Astrophysics
See the press release from HERA here:
https://www.sarao.ac.za/media-releases/telescope-built-by-karoo-locals-releases-data-to-astronomers-around-the-world/
Principal Investigator: Dr Mary-Jane Bopape
Institution Name: South African Weather Service
Active Member Count: 26
Allocation Start: 2021-08-04
Allocation End: 2022-02-09
Used Hours: 1642503
Project Name: Very Short Range Forecasting
Project Shortname: ERTH1022
Discipline Name: Earth Sciences
The South African Weather Service (SAWS) is collaborating with the Universities of Witwatersrand, Pretoria, NorthWest and Zululand to understand the Conformal Cubic Atmospheric Model (CCAM). The CCAM is a seamless model meaning it can be used across a range of timescales for weather forecasting to multi-decadal climate simulations. The model can also run as both a global model and a regional climate model. The model is being tested for numerical weather prediction purposes at SAWS because it provides an opportunity for human resources working on different timescales, to converge on one model, that can be improved locally. The model was found to be able to capture all high impact weather events, however it has some shortcomings such as underestimating the heaviest rainfall events and overestimating the number of small rainfall events. South African scientists are studying different aspects of the CCAM to improve it.
Principal Investigator: Prof Shazrene Mohamed
Institution Name: South African Astronomical Observatory
Active Member Count: 4
Allocation Start: 2021-07-29
Allocation End: 2022-02-03
Used Hours: 676737
Project Name: 3D rad-hydro simulations of the outflows of evolved stars
Project Shortname: ASTR0880
Discipline Name: Astrophysics
The computational stellar astrophysics group (PI. S. Mohamed) is currently based at the South African Astronomical Observatory and the University of Cape Town. The research uses supercomputers, powerful computational machines at the CHPC, to simulate the complex physics involved in the interactions of stars in the final phases of evolution. The interactions include exchanges with a nearby companion star or planet that can lead to the formation of spirals, disks, jets and even violent stellar explosions. By comparing the models with observations from powerful telescopes, these simulations help us understand how solar systems evolve and ultimately, the fate of our own star, the sun.
Principal Investigator: Dr Bubacarr Bah
Institution Name: African Institute for Mathematical Sciences
Active Member Count: 11
Allocation Start: 2021-08-04
Allocation End: 2022-02-21
Used Hours: 86196
Project Name: Large Scale Computations in Data Science
Project Shortname: CSCI0972
Discipline Name: Applied and Computational Mathematics
The Data Science research group at AIMS South Africa undertakes research on theoretical foundations of data science but also works on data science applications to solve real life problems.Example applications members of the group have been working on is the detection of fruits on images of fruit trees in orchards. This could be used for yield estimation and could be adapted for disease detection in these trees, etc. The image detection is done via deep learning models and the require huge computational resources, which are only provided by institutions like CHPC. One student working on the image detection project has submitted his MSc thesis for examination. Another students working on image classification using graphical neural networks also submitted his thesis for examination. Both students' work relied heavily on the computing resources of CHPC.
Principal Investigator: Prof Shahida Moosa
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-08-05
Allocation End: 2022-02-01
Used Hours: 46592
Project Name: Rare Disease Genomics
Project Shortname: HEAL1396
Discipline Name: Health Sciences
The Rare Disease Genomics in South Africa research group is based at Stellenbosch University. In late 2020, we established sub-Saharan Africa's first Undiagnosed Disease Programme (UDP). We have already completed testing on 100 patients, with undiagnosed rare diseases. Through the UDP more than half have received a diagnosis, which guides their treatment and management, and allows accurate genetic counselling. Our group's motto is "Until every african patient with a rare disease is diagnosed!". We have started in the Western Cape and are planning on expanding the UDP to beyond our geographical borders.
Principal Investigator: Prof Gerard Tromp
Institution Name: Stellenbosch University
Active Member Count: 8
Allocation Start: 2021-08-05
Allocation End: 2022-02-01
Used Hours: 137880
Project Name: South African Tuberculosis Bioinformatics Initiative
Project Shortname: CBBI0999
Discipline Name: Bioinformatics
The South African Tuberculosis Bioinformatics Initiative (SATBBI) performs an important role in tuberculosis research by supporting numerous research projects with data analysis and modeling. SATBBI also has an important educational and training mandate. The CHPC is a vital resource for these activities.
Principal Investigator: Prof Nithaya Chetty
Institution Name: University of Pretoria
Active Member Count: 4
Allocation Start: 2021-08-06
Allocation End: 2022-02-02
Used Hours: 716789
Project Name: Advanced Computational Resources for members of ASESMA
Project Shortname: MATS1050
Discipline Name: Material Science
The African School for Electronic Structure Methods and Applications (ASESMA) is a network of scientists in Africa who building up research in computational materials science. Our research interests for the computational and theoretical solid state physics group at the University of Pretoria mainly focus on the study of 2D materials. We are interested in the study of 2D hexagonal-like transition metal oxides (TMO) materials such as MoO2. It is predicted to be thermodynamically stable, however, a detailed study on the structural, electronic and dynamical properties of these new oxide material is still missing. Therefore, research about these properties is important and could bring them from the laboratory to commercial products. Recently (over the past two years) a 2-dimensional bilayer of zeolite has been experimentally synthesized by scientists in the USA. This is an area of promising new research for computational studies, and worth to be explored. We are mainly interested in the use of these 2D materials for catalysis such as CO catalytic oxidation. To achieve these goals, creating complex system such as introducing defects or adatoms will enhance the properties of the MoO2 and the 2D zeolite model. Our results provide detailed understanding and important information to the 2D materials community.
Principal Investigator: Prof Thuto Mosuang
Institution Name: University of Limpopo
Active Member Count: 10
Allocation Start: 2021-08-06
Allocation End: 2022-02-02
Used Hours: 17190
Project Name: Computational studies of various ultra-hard materials
Project Shortname: MATS0875
Discipline Name: Material Science
This is a sensors and energy materials group at the Department of Physics, University of Limpopo. The group has three (3) Doctoral, one (1) Masters, and one (1) Honours students. Computationally the research projects investigate nanomaterials like copper selenide, gallium nitride, gallium arsenide, graphene oxide, boron nitride, gold and silver nanoparticles. The gold and silver nanoparticles are being research for possible toxicity/non-toxicity when ingested in human tissues. Specifically, electronic, structural, optical, thermodynamic properties are investigated to enhance semi-conductivity. These properties are then mapped with experimental properties for possible gas sensing and energy materials. DL_POLY software through the CHPC is being utilised in the structural, dynamical, and thermodynamic properties. The exciting code also through the CHPC is being utilised in the electronic and excited properties of the materials. Now lately, Materials Studio also through CHPC is being used to structural and dynamic properties of gold and silver nanoparticles.
Principal Investigator: Dr Malebogo Legodi
Institution Name: University of Venda
Active Member Count: 2
Allocation Start: 2021-08-10
Allocation End: 2022-02-06
Used Hours: 75120
Project Name: Synthesis and characterization of Na2Fe2(MoO4)3 solid solutions for use in sodium ion batteries
Project Shortname: CHEM1268
Discipline Name: Chemistry
Name of Research Group: Inorganic Energy Storage System, Department of Chemistry, University of Venda; Nature of Work: The research program involves metal doping of Na2Fe2(SO4)3 at Na and Fe sites using alkali/alkali-earth and transition metals, respectively. The simulations will predict the most stable form and structure of the resulting products, thus guiding the synthesis process. It is envisioned that the doping process will improve electrochemical properties of the resulting cathode material. Why the work is done: The work is carried out as contribution to the broad solution of energy management. As the fossil fuels are gradually replaced by renewable energy sources, batteries become even more relevant for storage of renewable or green energy. The products from this study can be used as cathode material in the future sodium ion batteries. Process: The method used for synthesis of metal-doped cathode material is glycine nitrate soft combustion method. It involves the mixing in stoichiometric amounts, vapourization at 100 oC to form gel. The gel is then calcined at 600 oC to give black powder as product. Organic electrolyte materials that will be compatible with proposed cathode materials will also be investigated by calculating molecular energies, HOMO and LUMO energies. Progress The input have been developed for simulation and few runs carried out.
Principal Investigator: Prof Amanda Rousseau
Institution Name: University of the Witwatersrand
Active Member Count: 3
Allocation Start: 2021-08-10
Allocation End: 2022-02-06
Used Hours: 1435
Project Name: Antifolates and antimalarial kinase inhibitors
Project Shortname: CHEM1402
Discipline Name: Chemistry
Prof A Rousseau and collaborators in the School of Chemistry at the University of the Witwatersrand are working on the design and synthesis of inhibitors of folate metabolism. Folate inhibitors, known as antifolates, and in particular inhibitors of the enzyme dihydrofolate reductase (DHFR), find application as treatments in a range of disease classes. We are focusing on the design of antifolates for the treatment of malaria, a disease that is prevalent in Africa and caused by parasites of the genus Plasmodium, with Plasmodium falciparum being responsible for the most malaria fatalities worldwide. In order to do so, we are utilising molecular modelling as a tool to assist with the design of inhibitors by using crystal structures of both wild type and mutant forms of the enzyme DHFR present in the P. falciparum parasite (PfDHFR). Molecular modelling using Schrodinger software allows us to computationally assess the potential of compounds of interest to bind in the active site of the PfDHFR enzyme. This helps us to determine which compounds to synthesise, and, once the compounds have been assessed for biological activity in vitro, can be used to optimise the biological activity against the target. The time taken to complete the computational assessment (in silico screening of inhibitors) is greatly reduced through the use of the CHPC, as multiple jobs can be submitted at once. Furthermore, more complex dockings which allow for movement of the amino acids within the protein (flexible docking protocols) can be readily accomplished. We have assessed two series of compounds containing a pyrimidine core in silico, synthesised a small set of analogues in each series and currently await the results of in vitro assessment of biological activity of these compounds.
Principal Investigator: Dr Chika Nnadozie
Institution Name: Rhodes University
Active Member Count: 1
Allocation Start: 2021-08-11
Allocation End: 2022-02-16
Used Hours: 21243
Project Name: Campylobacteroisis
Project Shortname: CBBI1446
Discipline Name: Bioinformatics
This research group is hosted in the institute for water research (IWR), Rhodes University, Grahamstown, Eastern Cape. IWR is a renowned research and innovation center specialiszing in environmental water quality reasearches. The research group interest include researches in the areas of computational and molecular microbiology, risk assessment and public health, microbial water quality and ecology as well as integrative health research, linking laboratory research to policy outcomes and practice. This project focuses on the sources, exposure rooutes and potential risk of Campylobacter infections from selected rivers in the Easern Cape. Campylobacter spp cause campylobacteroisis, a chronic enteric infection in humans. They are responsible for approximately 500 million cases of gastroenteritis per year globally. Contact and consumption water contaminated with feaces is a major risk factor. South African provinces including Eastern Cape Province (ECP) rely on available surface water bodies such as rivers for agricultural, domestic and recreational purposes. However, South African provinces is faced with problem of sewage treatment crisis, poor operation and design of wastewater treatment plant, which could lead to suboptimal wastewater treatment of effluents before discharge into nearby rivers. The problem is compounded by high number of pipeline leakages which results in inflow of faeces-associated pathogens into the rivers, shedding and excreta from livestock that come to the river to drink water as well as run-offs from nearby farms. All these are potential sources of Campylobacter spp. to the Kowie River.
Principal Investigator: Dr Abdulrafiu Raji
Institution Name: University of South Africa
Active Member Count: 13
Allocation Start: 2021-08-11
Allocation End: 2022-02-21
Used Hours: 4270781
Project Name: Structural, electronic, magnetic properties and anisotropy energy in some metallic and non-metallic heterostructures
Project Shortname: MATS0988
Discipline Name: Material Science
The research work is a collaboration between Dr. Raji (Physics Department, UNISA, South Africa) and the group of Dr. Brice Rodrigue Malonda (Marien Ngouabi University, Republic of Congo). The focus of the research is numerical studies of electronic, optical, transport and magnetic properties of selected two-dimensional (2D) and three-dimensional (3D) solid materials for potential applications in high-capacity data storage, catalysis and renewable energy. The study employs density-functional theory (DFT) investigate the atomic-level properties of these materials. We aim to modify the pristine properties of selected solids through simple defect engineering of their structure, alloying and interface engineering.
The last two decades have witnessed efforts at discovering materials with novel properties for specific applications such as in electronic and magnetic devices. The discovery of new materials are sometimes accompanied by previously unknown physical, chemical or electronic processes which necessitate the use of advanced theoretical approaches to gain deeper understanding of these processes. One of the aims of our research therefore, is to discover novel materials via computational methods and to exploit the unique properties of these materials for potential technological applications. Ultimately, our research is theory-led discovery of novel materials.
The numerical implementation of DFT is computational intensive, requiring several computational hours, large data storage and memory requirements, beyond the capacity of ordinary table-top computers. Therefore, high-capacity computer clusters such as the one provided by the Center for High Performance Computing (CHPC) is sine-qua-non for the research. There are about 10 postgraduate students of various nationalities working in various aspects of the project. This research, as well as the CHPC, has enabled collaboration between South Africa based academic researcher and colleagues in Congo, Mexico and Italy. There have been research outputs in form of publications, and the scope of the project will guarantee regular postgraduate students training who are able to undertake cutting-edge research and contribute to scientific development of South Africa, DRC and the Africa continent.
Principal Investigator: Dr John Mack
Institution Name: Rhodes University
Active Member Count: 2
Allocation Start: 2021-08-11
Allocation End: 2022-02-07
Used Hours: 15366
Project Name: Rational design of phthalocyanine, porphyrin and BODIPY dyes
Project Shortname: CHEM0796
Discipline Name: Chemistry
The Institute for Nanotechnology Innovation (INI) at Rhodes University under director Prof. Tebello Nyokong carries out research related to the use of molecular dyes in biomedical applications, such as cancer treatment through photodynamic therapy, the use of antimicrobial photodynamic chemotherapy for treating hospital superbugs and as sensors for ions that are harmful to human health.
Other applications of interest include wastewater treatment and the development of optical limiters to protect human vision,such as in aviation safety in protecting pilots from the irresponsible use of laser pointers during runway approaches. It is important to understand trends in the electronic and optical properties that make the molecular dyes suitable for these particular applications. The resources of the Centre for High Performance computing help to facilitate this by making it possible to carry out molecular modelling calculations that can be used to predict how changes to the structures of the molecular dyes will modify their properties. The flexibility that CHPC provides where memory allocations are concerned is often vital.
The INI currently has three staff members, fifteen PhD and six MSc students, and three postdoctoral fellows, and usually publishes over forty peer-reviewed publications per year. This rose to a high of sixty-five in 2017 with ten involving the use of CHPC resources.
Principal Investigator: Dr Brigitte Glanzmann
Institution Name: Stellenbosch University
Active Member Count: 20
Allocation Start: 2021-08-11
Allocation End: 2022-02-07
Used Hours: 319287
Project Name: SAMRC Precision Medicine African Genomics Centre
Project Shortname: CBBI1195
Discipline Name: Bioinformatics
The South African Medical Research Council's Genomics Centre provided the first whole genome sequencing facility to Africa. Since our inception in 2019, we have sequenced numerous samples, ranging from humans to lions and tigers, using whole genome, exome and transcriptome sequencing. All of the data generated is processed locally using the CHPC, an invaluable resource on which the Genomics Centre is reliant.
Principal Investigator: Prof Nelishia Pillay
Institution Name: University of Pretoria
Active Member Count: 12
Allocation Start: 2021-08-11
Allocation End: 2022-02-21
Used Hours: 162913
Project Name: Nature Inspired Computing Optimization
Project Shortname: CSCI0806
Discipline Name: Computer Science
The NICOG (Nature Inspired Computing Optimization Group), based at the University of Pretoria, employs machine learning and optimization techniques, taking an analogy from nature, to solve real world problems to attain the sustainable development goals defined by the United Nations. This research is solving problems in the health sector, e.g. disease prediction, agriculture, e.g. disease, crop modelling and streamflow prediction and industry, e.g. logistics, scheduling and classification problems. As we move into the fourth industrial revolution, machine learning is playing a major role in solving these problems. The algorithms used to solve these problems are computationally expensive and hence high performance computing is needed for the implementation of the algorithms. We are at point where we need to make these approaches easily accessible to non-experts. Hence, one of the current foci of NICOG is the automated design of machine learning techniques to produce off-the-shelf tools for non-experts. Such automated design, which is essentially an optimization problem, is also computationally expensive and thus also requires high performance computing.
Principal Investigator: Prof Thulani Makhalanyane
Institution Name: Stellenbosch University
Active Member Count: 13
Allocation Start: 2021-08-11
Allocation End: 2022-02-07
Used Hours: 48422
Project Name: Metagenomics of extreme environments
Project Shortname: CBBI1023
Discipline Name: Bioinformatics
Although a major part of our work focuses on investigating the contribution of microbial communities (bacteria, archaea and fungi) towards ecosystem processes in the ocean, some aspects of our research include assessing freshwater systems and how they may impact human health. Some of these systems which include rivers and lakes are impacted by unmaintained sewage systems and other non-point source pollution sources such as stormwater runoffs. Ongoing related studies have indicated that such pollutants carry pathogens that may contaminate drinking water supplies, which thus lead to outbreaks that regularly affect underdeveloped communities. Therefore, the overall goal of the work we do within the Microbiomics Research Group which links to freshwater systems is to mainly define point sources of pollution and develop new approaches that can be utilized with ease to prevent outbreaks, protect rivers, lakes, human health and help inform policy makers. To reach our goal we need to use CHPC facilities to mine freshwater associated data for key gene targets.
Principal Investigator: Prof Tobi Louw
Institution Name: Stellenbosch University
Active Member Count: 8
Allocation Start: 2021-08-11
Allocation End: 2022-02-07
Used Hours: 87380
Project Name: SU Process monitoring, modelling and control
Project Shortname: CSCI1370
Discipline Name: Applied and Computational Mathematics
The SU Process monitoring, modelling and control research group aims to improve industrial chemical- and minerals-processing using simulation based studies. Our interest ranges from the application of machine learning to improve process efficiencies by detecting faults as (or before) they occur, to the use of sophisticated fluid dynamics models to understand how chemical reactors work. Currently, we are busy with two exciting projects: one investigating a digester for the pulp and paper industry, another considering the complicated flows in a novel bioreactor.
Principal Investigator: Prof Eric K. K. Abavare
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 5
Allocation Start: 2021-08-11
Allocation End: 2022-02-07
Used Hours: 306332
Project Name: Atomic and Electronic structures of semiconductor interface systems
Project Shortname: MATS1159
Discipline Name: Physics
The Frontier Science Group (FSG) in the Department of Physics, Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, Ghana is proud to be associated with the Centre for High Performance Computing (CHPC) in South African. We investigate atomistic and electronic properties of light materials using quantum mechanical approaches based on first principles calculations in the framework of density functional theory (DFT).
Our researches basically focus on the study of interface and magnetic materials. All electronic devices and related components in atomic level are interfaced. We study these interfacial morphologies and how they affect the overall performance of device applications.
Experimentally, several important semiconductor materials are growth based, as a consequence complete understanding of how these growths are achieved are certainly important for device design and application.
Specifically, our interested lies with interface of silicon carbide (SiC) growth on silicon (Si) and similarly gallium nitride (GaN) on silicon for power electronic materials, for the magnetic materials, we concentrate on doping of the -phen materials; that is the doping of the single layer of group IV materials.
These materials are important for power applications and especially crucial in hostile environment. Nevertheless, they exist in small crystal sizes and to obtain large size for device application, researchers are employing all techniques to grow them on silicon since silicon technology is well established and could be easily integrated. The key issue, which troubles the perfect layer-by-layer matching growth, is the lattice mismatch existing between the materials themselves with silicon.
By careful special crystal orientation, the materials involve above could be made to match and perfect growth achieved, which means large crystal sizes could be formed. However, the matching mechanism is not well understood and this needs computational approaches for clarification and here CHPC comes to fore in this search. The CHPC is an indispensable tool to help our group pushes forward the frontiers of this area for mankind in our quest for the pursuit of scientific achievement and better human life through technology.
Principal Investigator: Dr Kathryn Wicht
Institution Name: University of Cape Town
Active Member Count: 3
Allocation Start: 2021-08-11
Allocation End: 2022-02-07
Used Hours: 22312
Project Name: New tools for antimalarial target identification
Project Shortname: HEAL1448
Discipline Name: Chemistry
Who: Bioinorganic chemistry, UCT
What: In silico screening against Plasmodium falciparum targets
Why? Identify novel scaffolds for medicinal chemistry optimisation
How? Docking and MD simulations
Progress: In silico hits have been identified that show activity against P. falciparum at 1uM.
Principal Investigator: Prof Michelle Kuttel
Institution Name: University of Cape Town
Active Member Count: 3
Allocation Start: 2021-08-12
Allocation End: 2022-02-08
Used Hours: 78502
Project Name: Molecular modelling of microbial polysaccharides and glycoconjugates
Project Shortname: CHEM1242
Discipline Name: Chemistry
The project is run by Prof. Michelle Kuttel of the Department of Computer Science at the University of Cape Town. We focus on molecular modelling to inform the development of more effective drugs and vaccines against infectious disease. This work aims to provide answers to important questions on the shape and interactions of pharmacologically active that are difficult to obtain experimentally. We run and analyse atomic simulations of molecules of interest on the CHPC computers. We have a number of successful publications on this work.
Principal Investigator: Prof Regina Maphanga
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 7
Allocation Start: 2021-08-12
Allocation End: 2022-02-08
Used Hours: 251258
Project Name: Energy Storage Materials
Project Shortname: MATS0919
Discipline Name: Material Science
The Design and Optimisation group is a group under Operational Intelligence Impact Area in the Next Generation Enterprises and Institutions cluster of CSIR and uses multiscale methods to develop novel energy storage materials with desired properties and enhanced properties for existing materials. The envisaged outcome of this project is to develop energy storage materials models with improved performance and consistency, which will be cheaper and environmentally benign. The ever-increasing global energy needs and depleting fossil-fuel resources demand the pursuit of sustainable energy alternatives, including renewable sources to replace carbon intensive energy sources. Sustainable and renewable energy is considered to be the most effective way to minimize CO2 emissions. Hence, finding sustainable energy storage technologies is vital for optimally harnessing the renewable energy.
Computational methodologies are very effective in predicting material-structure-property correlations. In this project simulation methods and models based on parallel computing are developed to probe energy materials properties. Centre for High Performance Computing (CHPC) resources are used to develop these parallel computing methods and algorithms for large and complex material models. Thus, CHPC resources provide a platform to simulate the evolution of material properties at a wide range of external conditions that are not accessible experimentally and fast track acquisition of results. The major of aim of this project is to create a predictive, reliable and robust set of models of materials for energy storage across the materials modelling length scale, leading to integrated multi-scale capability.
Principal Investigator: Prof Johan W. Joubert
Institution Name: University of Pretoria
Active Member Count: 3
Allocation Start: 2021-08-12
Allocation End: 2022-02-08
Used Hours: 245595
Project Name: Mobility modelling using vehicle telematics
Project Shortname: MECH0803
Discipline Name: Other
Government (all spheres) is tasked with the challenge of deciding where to invest in infrastructure (for housing/transport/economic development etc). The real hard part is to balance two questions. Firstly, "who gets the benefit of infrastructure?" and, secondly, "who PAYS for those benefits?". In our research programme, we develop agent-based transport models that are able to depict and imitate the complex, multi-modal and large-scale transport scenarios in a South African context. Each individual - from the Metrorail commuter to the private car owner, to the minibus taxi driver, to the commercial vehicle delivering your goodies at the Woolies Foods - are represented in such a transport model. The goal of our ongoing research is to assist government and industry make better informed, evidence-based decisions about the intended, and unintended consequences of specific transport interventions. Since every individual is modelled the computational burden is substantial and the CHPC allows us to overcome the data preparation and simulation challenges of executing the large-scale, MATSim models.
Principal Investigator: Prof Paulette Bloomer
Institution Name: University of Pretoria
Active Member Count: 7
Allocation Start: 2021-08-16
Allocation End: 2022-02-12
Used Hours: 37055
Project Name: Molecular Ecology and Evolution Programme
Project Shortname: CBBI1030
Discipline Name: Bioinformatics
The Molecular Ecology and Evolution Programme (MEEP) is a research group in the Division of Genetics, Department of Biochemistry, Genetics and Microbiology at the University of Pretoria. We investigate the evolutionary history and current genetic diversity of many vertebrate species (marine and terrestrial), using genetics and genomics tools, and try to link the patterns we see with ecological and human-mediated events. We currently have several projects (Cape buffalo and different birds including falcons and Houbara) that make use of the CHPC resources, with additional projects in the pipeline.
Genomics of Cape Buffalo is a hot topic and is of interest to the general public and wildlife ranchers in particular. Cape buffalo are bred on private game ranchers for a one or a few economically important traits, such as horn length and body size. The effects of this artificial selection, as well as breeding South African buffalo with East African buffalo, is unknown. We can use genomics to evaluate the risks and benefits of these practices for the short and long term (evolutionary) conservation prospects of the species
The genetic work on the bowhead whale uses empirical and simulation data to investigate the effects that different Nearctic human cultures exploiting the whale had on the abundance of the population. This means that genetic data may have more resolution than previously thought and is sufficient to detect subtle changes in population size in relation to exploitation and climate change. This is of great importance as there are many applications. We believe that the work will have sufficient impact to be of interest to the general public in South Africa and internationally.
The development of a new molecular clock calibration method for tropical reef species makes use of simulations for validation. This calibration enables genetic time to be related to chronological time. This method may then be applied to empirical data to identify driving factors for increases in population size of reef dependent species in the tropics.
Several research projects on birds have benefited from the CHPC over the last six months. These include falcons and houbara that are important for conservation and management perspectives.
Principal Investigator: Dr Jaco Badenhorst
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 3
Allocation Start: 2021-08-16
Allocation End: 2022-02-22
Used Hours: 20871
Project Name: Automatic transcription of broadcast data
Project Shortname: CSIR0978
Discipline Name: Other
The Voice Computing (VC) Research Group at the CSIR Nextgen Enterprises and Institutions (NGEI) cluster develop speech and language-related technologies for the South African context to enhance access to information and communication. Given the multilingual nature of South Africa, human language technologies (HLTs) can make information and services accessible to a larger proportion of the South African population in a sustainable way, by reaching people in remote locations, people who are not necessarily trained in using technology, or people who are not fluent in English. HLTs can also provide access to information to people with disabilities. HLT can support language diversity and providing of information in multiple languages in an affordable and equitable fashion. The development of all 11 official languages is a national priority, which requires significant attention to HLT in all of these languages. HLT can be of significant economic importance – both by empowering citizens of the country to work more productively, and by forming the core of an exportable set of technologies (since such technologies are currently not available to most of the developing world).
Principal Investigator: Dr Samuel Iwarere
Institution Name: University of Pretoria
Active Member Count: 2
Allocation Start: 2021-08-16
Allocation End: 2022-02-12
Used Hours: 114239
Project Name: Understanding Plasma-Liquid Interaction
Project Shortname: MECH1107
Discipline Name: Chemical Engineering
The research is being carried out under the guidance of the Thermodynamics Research Group based in the School of Chemical Engineering at the University of KwaZulu-Natal. The understanding of plasma-liquid interactions is of prime importance in the context of environmental remediation and wastewater treatment applications. Plasma treatment of wastewater involves complex chemical and physical phenomena, which makes the comprehensive modelling of phenomena such as gas phase reaction, turbulence and heat transfer a difficult task. In order to gain insight into the complex processes associated with the use of plasma wastewater treatment, Computational Fluid Dynamics (CFD) will be used to study the relevant chemistry and fluid dynamic effects at the plasma-liquid interface. The Centre for High Performance Computing (CHPC) allows for calculation within realistic time frames, allowing the researchers to determine the accuracy of the simulations encountered in plasma physics. Simulation of the plasma arc is progressing well with the researchers working in conjunction with other scientists in the plasma field.
Principal Investigator: Prof Gert Kruger
Institution Name: University of KwaZulu-Natal
Active Member Count: 17
Allocation Start: 2021-08-16
Allocation End: 2022-02-12
Used Hours: 280412
Project Name: CPRU molecular modeling
Project Shortname: HEAL0839
Discipline Name: Health Sciences
Who) The Catalysis and Peptide Research Unit at UKZN (http://cpru.ukzn.ac.za/Homepage.aspx) is currently working on the mechanism of action on the HIV PR with respect to the natural substrates. (What) We use an quantum chemical/molecular mechanics hybrid (Oniom) approach that gives us high level results on the thermodnamics and kinetics of the reaction. Our model enables us to study application of both electronics and sterics to change the bond strenght and binding energies of new potential inhibitors. The same approach is followed for transpeptidases, essential enzymes for the inhibition of TB. (Why) HIV is a major health threat in SA and new drugs is required due to the constant development of drug resistance. Similarly, TB and TB co-infection with HIV is a mojor health issue. (How) Our theoretical studies will allow us to test a computational model that correctly describes the interactions between the respective enzymes and existing drugs. We are improving our computational model to the extent where are now in a position to start proposing new potential drug leads for synthesis. This can only be done with large computational resources provided by the CHPC.
Principal Investigator: Prof Tien-Chien Jen
Institution Name: University of Johannesburg
Active Member Count: 20
Allocation Start: 2021-08-17
Allocation End: 2022-02-23
Used Hours: 1183561
Project Name: Developing an Innovative Metal Matrix Composite Membrane for Hydrogen Purification
Project Shortname: MATS0991
Discipline Name: Material Science
Hydrogen is an excellent energy carrier because of its high enthalpy of combustion with less environmental impact. Hydrogen fuel can be directly used in cars, ships, aircraft, and other transportation equipment. Although there are many problems to be resolved for implementation of hydrogen technology, it is generally believed that hydrogen will become a key energy carrier in this century. Until sufficient quantities of hydrogen can be derived from renewable energy sources, it will have to come from hydrogen separation using existing fuels (e.g., diesel, natural gas) as the feedstock. Hydrogen separation units can generate highly purified hydrogen from the feedstock gas mixtures, however, a certain amount of CO and other chemicals will make it into the hydrogen stream and cause catalyst failure on the fuel cell electrode. This significantly deteriorates the activity of the catalyst layer, particularly for Proton exchange membrane (PEM) fuel cells. This deterioration greatly reduces the efficiency of power generation and the life cycle of the fuel cell. Overcoming this obstacle is necessary for the wider use of highly efficient fuel cells. The objective of this proposal is to develop a new composite metal membrane (CMM) that can achieve extremely high flux and selectivity of the hydrogen separation and achieve the lifetime goal set by the U.S. department of energy (DOE). To accomplish these goals, a research team formed by the University of Johannesburg (UJ), Council for Scientific and Industrial Research (CSIR) and Hydrogen South Africa (HySA) will combine two major technologies: Electrostatic-Force-Assisted Cold Gas Dynamic Spraying (ECGDS) and Laser-Assisted Manufacturing (LAM) to develop this new type of membrane. The proposed new membrane will eliminate the hydrogen embrittlement, catalyst failure, and avoid the difficulty of forming a defect-free, ultrathin Pd film on the substrate surface. The proposed CMM will exhibit superior reliability and can be produced with a relatively low-cost and scalable process. Along with the development process, numerous numerical methods are utilized to simulate the material properties and behaviour as well as the optimization fabrication procedure. To accomplish these task HPC services from CHPC plays a crucial role in understanding the numerous unknown phenomena's. This service enables the research to be simulated from feature scale to reactor scale .Note that this project is currently supported by the NRF for 2019-2022.
Principal Investigator: Dr Samson Khene
Institution Name: Rhodes University
Active Member Count: 3
Allocation Start: 2021-08-17
Allocation End: 2022-02-13
Used Hours: 2141
Project Name: Synthesis, Spectroscopy and Nonlinear Optical Properties of Phthalocyanines
Project Shortname: CHEM0975
Discipline Name: Chemistry
Many of today's optical sensors are based on optical materials in which non-linear optics may play a role. Defects in the cable can have an effect of reducing the light (which carries information) in an optical fiber or cable. This phenomenon is also present in a variety of non-linear active materials or molecules. This nonlinear reduction of light phenomenon can be used as a sensing method for detection and correction of signals that are sent in cables. The aim of our research (at Rhodes University in the Chemistry department) is to make and study molecules which will help improve signal transmission using a nonlinear process. This work will not be possible if it was not supported by CHPC facilities.
Principal Investigator: Prof Daniel Joubert
Institution Name: University of the Witwatersrand
Active Member Count: 6
Allocation Start: 2021-08-17
Allocation End: 2022-02-13
Used Hours: 286215
Project Name: Energy Materials: Numerical explorations
Project Shortname: MATS0800
Discipline Name: Physics
Imagine designing a material atom by atom. Imagine analysing the properties of tomorrow's novel materials before they exist. Computational materials science plays an important role at every of level of the design and engineering of new materials. The rapid advances in computer processing power and memory has given virtual, fundamental, materials design a boost. The first problem faced in designing a virtual material, a material that has not yet been made, is to find the stable configurations in which the atoms in the mix will settle into. The designer must find the configuration of electrons and nuclei, the building blocks of atoms, which result from the interactions among a number of particles that, even for a small piece of material, exceeds the number of particles of sand on all the beaches in the world. This daunting task has a simple solution. Instead of finding the configuration of the real material, the problem is solved for a fictitious material where the constituent particles do not interact. All that is necessary is to find the particle density distribution of the fictitious material, which by design, is the same as that of the real material. A Noble prize was awarded for this idea to the chemists Walter Kohn and John Pople in 1998.
A group of postgraduate students at the University of the Witwatersrand use the computer power at the national Centre for High Performance Computing to conduct virtual experiments on existing and novel materials to examine their potential as energy harvesters. They examine the potential of a selection of materials that can be used as cheap components in solar cells, materials that can generate electricity from waste heat and materials that can be used to split water molecules to extract hydrogen for energy production.
Principal Investigator: Dr Samuel Atarah
Institution Name: University of Ghana
Active Member Count: 1
Allocation Start: 2021-08-19
Allocation End: 2022-02-15
Used Hours: 35036
Project Name: Ab initio studies of electronic and magnetic properties of selected elements
Project Shortname: MATS1162
Discipline Name: Physics
In the Condensed Matter Physics group in the Department of Physics at the University of Ghana. Part of our recent research effort is to study perovskite material with potential application to solar energy harvesting. We focus on their physical and electronic properties. Our work is motivated by their potential applications to clean and renewable energy. A typical process involves setting up a basic unit or cell of the perovskite of structure A2BB'X6; making a self consistent quantum mechanical calculation of its minimum total energy, its bands structure, and density of states. Thereafter the bands structure is recalculated using another material A2BB'X'6 (where A is a metal, B is Ag and B' is In whilst X and X' are two different halide atoms). The changes in band structure or density of state is studied as a function of the halide atoms. Initial results are still being analyzed and further investigations planned. With the support of the CHPC computing power, the project is seeing real progress.
Principal Investigator: Dr Thishana Singh
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-08-20
Allocation End: 2022-02-16
Used Hours: 8699
Project Name: Combined kinetics,quantum-chemical investigation of reaction mechanisms involving catalysts
Project Shortname: CHEM0821
Discipline Name: Chemistry
The TS Computational Chemistry Research Group (University of Kwazulu-Natal, School of Chemistry and Physics) currently, uses computational chemistry as a tool in the field of Green Chemistry to compare the chemical reactivity properties of bioactive molecules for drug delivery. Conceptual DFT which is often referred to as 'computational nutraceutics' is a concept used to predict the molecular structure, spectroscopy, and chemical reactivity of nutraceuticals (food or part of a food that provides medical/health benefits, including the prevention/treatment of diseases) by means of computational chemistry and molecular modelling. Using this methodology, the aim is to perform a combined electronic and structure investigation to understand enzyme-inhibitor interactions that can aid in the synthesis and strategic design of novel drugs for the treatment of various infectious diseases. The CHPC provides the facilities: hardware, in the form of the Lengau cluster and software programs such as Gaussian16, amongst many others.
Principal Investigator: Prof Robinson Musembi
Institution Name: University of Nairobi
Active Member Count: 7
Allocation Start: 2021-08-20
Allocation End: 2022-02-16
Used Hours: 101198
Project Name: CMCG-UoN
Project Shortname: MATS1321
Discipline Name: Material Science
The name of the group is MATS1321, Condensed Matter and Computational Physics Group domiciled at the University of Nairobi. The group is a subset of a larger group of Solid-state Physicists though the other group members are experimentalists. The group focus is multifaceted in diverse areas, some members are working on molecular dynamics, while the other members are working on density functional theory for material which form crystal structures which are regular. The importance of the research is that some of our students are researching on drug discovery related research which requires computing facilities to accelerate the rate of doing the work and optimise the values such that the experimentalist will spend less time in coming up with the final product. The other research group members are working on energy materials, materials for solar cells or materials which produce electrical current when subjected to certain conditions like change in temperature. Some of the materials may have superconducting properties which are important in manufacturing equipment like the magnetic resonance imaging.
Principal Investigator: Dr Rose Modiba
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 6
Allocation Start: 2021-08-20
Allocation End: 2022-02-23
Used Hours: 377904
Project Name: Computational modelling of light metals
Project Shortname: MATS1097
Discipline Name: Material Science
The Advanced Materials Engineering group under the Manufacturing cluster at CSIR who are working on the stability of the materials when alloyed with other transition metals would like to thank CHPC for allowing us to use their facilities. The systematic search for novel materials is hardly achievable only via experimental discovery due to the very large combinatorial space spanned by the crystal structures and chemical compositions for compounds. This difficulty is also connected with eventually high and unstable costs of raw materials on the world market and the time-consuming procedures of synthesis and characterization. It is of great importance for the theoretical search for materials to gain detailed insight into the most relevant physical mechanisms which determine the intrinsic properties of the materials. The CHCP is a very good platform for anyone who would like to do simulation on their materials. Thank you CHPC for the workshops provided to equip us with the most relevant and recent technologies to be able to understand our materials.
Principal Investigator: Prof Francois Engelbrecht
Institution Name: University of the Witwatersrand
Active Member Count: 3
Allocation Start: 2021-08-23
Allocation End: 2022-02-25
Used Hours: 800930
Project Name: WITS Climate Modelling
Project Shortname: ERTH1200
Discipline Name: Earth Sciences
The Global Change Institute (GCI) of Wits, using the CHPC Lengau clustre, performed and innovative regional climate model experiment over East Africa. The simulations are the first where a regional climate model was coupled to a fully dynamic model of circulation in Lake Malawi, including a full representation of land-atmosphere fluxes over Lake Malawi. The simulations realistically simulate the development of a pronounced lake-breeze on the western shore of Lake Malawi. A journal publication on the simulations is being prepared by the GCI climate modellers, and is expected to provide new insights into the pronounced rainfall maxima observed over the lake.
Principal Investigator: Dr Annerine Roos
Institution Name: University of Cape Town
Active Member Count: 7
Allocation Start: 2021-08-23
Allocation End: 2022-02-19
Used Hours: 24850
Project Name: DCHS and MRI studies
Project Shortname: HEAL1267
Discipline Name: Health Sciences
We use CHPC sources to look at brain imaging data in various adult psychiatric disorders to better understand, for instance, OCD and depression. We collaborate with international groups to bring together datasets of thousands of people on these disorders so that we can better understand how the brain may be different. The people participating in these studies are of all walks of life and different nations. The ultimate aim is to then find better treatments for these disorders.
We also look at data of the Drakenstein Child Health Study to investigate determinants of early child development in a cohort of >1000 mother-child dyads of the Drakenstein district in the Western Cape of South Africa. The communities participating in the study represent a low-middle income country setting that is characterised by high substance abuse, poverty and violence. The Universities of Cape Town, Stellenbosch and Los Angeles, California work together in these studies. We aim to assess cognitive, behavioural and neural outcomes in children from birth to 6 years of age, to identify vulnerable groups and key periods for intervention that may optimise development. Evidence shows the earlier the interventions, the better the outcomes.
For this study, we use CHPC resources to analyse brain imaging data including structural data, of children with prenatal exposure to maternal depression and substances (e.g. alcohol and tobacco smoking). Prenatal exposure to maternal depression and/or substances may adversely affect the brain's connections, that may present as impaired cognitive and behavioural development in offspring.
Principal Investigator: Prof Matthew Adeleke
Institution Name: University of KwaZulu-Natal
Active Member Count: 11
Allocation Start: 2021-08-23
Allocation End: 2022-02-19
Used Hours: 105713
Project Name: Computational and Evolutionary Genomics
Project Shortname: CBBI1231
Discipline Name: Bioinformatics
The Quantitative and Computational Genomics is using genomics approach to under how probiotics modulates the rumen microbiome for improved health and productivity of small stock in South Africa. Furthermore, our group seeks to understand genetic and antigenic diversity of coccidian parasites for designing anticcodial vaccines and in general control measures for improved animal welfare and increased productivity. Part of the study's focus is to utilize computational approach to understand selection signature and explore comparative genomics to vaccine candidates. Samples are usually collected for DNA extraction. Genetic markers of interest are then applified and the product is then sent for sequencing. The CHPC platform is then utilized for analyzing Next Generation Sequencing data to provide interpretation.
Principal Investigator: Prof Adrienne Edkins
Institution Name: Rhodes University
Active Member Count: 1
Allocation Start: 2021-08-23
Allocation End: 2022-02-19
Used Hours: 365258
Project Name: Molecular and Cellular Biology of the Eukaryotic Stress Response
Project Shortname: CBBI1405
Discipline Name: Health Sciences
The Biomedical Biotechnology Research Unit (BioBRU) is a biomedical research unit at the Department of Biochemistry and Microbiology at Rhodes University. We are a collaborative research unit focused on understanding the structure and function of the cellular stress response. The current study identifies new drug targets in Mycobacterium tuberculosis, the causative agent of tuberculosis (TB) disease. We are using in silico modelling to study the native and mutant versions of clinically relevant TB proteins to examine if such mutations destabilize the protein. We hypothesize that our proposed new drug targets are required for stabliising mutant proteins associated with TB drug resistance. Considering that this analysis is computationally resource-demanding, the CHPC platform is invaluable to our study.
Principal Investigator: Prof Mpho Sithole
Institution Name: Sefaku Makgatho Health Sciences University
Active Member Count: 7
Allocation Start: 2021-08-24
Allocation End: 2022-02-20
Used Hours: 27706
Project Name: Computational modeling of titanium based alloys
Project Shortname: MATS1228
Discipline Name: Physics
The research group is based at Sefako Makgatho Health Sciences University. This is the 3rd year of the group. The group is collaborating with colleagues from CSIR Pretoria. The research projects are based on Titanium alloys [Ti-Mo-Zr] for biomaterial used as bone plates and orthopaedic implants and permanent magnets [Nd-Fe-B] for electronic appliances, generators and electric cars. The project employ the First Principle Method using CASTEP in Material Studio. The calculations are based on DFT. The group rely on CHPC for handling huge data and for high speed data calculations.
Principal Investigator: Mrs Chantel Niebuhr
Institution Name: University of Pretoria
Active Member Count: 2
Allocation Start: 2021-08-25
Allocation End: 2022-02-21
Used Hours: 257293
Project Name: Hydrokinetic turbine analysis
Project Shortname: MECH1174
Discipline Name: Other
The hydropower research group at the University of Pretoria focus on the development and optimisation of small hydropower schemes in South Africa. Currently computational fluid dynamics (CFD) modelling is being used to develop an in depth understanding of the hydrodynamic effects (wake and backwater build up effects) from a hydrokinetic device placed in inland flow. Understanding of these effects is necessary to minimize costs and unfavourable conditions (such as underperforming turbines, channel scour, channel overflow etc.). Due to increasing computational abilities CFD modelling is becoming more and more possible in this complex flow field. Using Reynolds stress RANS models, digital twins of laboratory tests were modelled and validated with experimental results, proving the usefulness of the models and allowing determination of the best functioning models within this type of analysis. Currently the models are being used to vary operational conditions and analysis the changes in the flow field. This is then used to determine simplified analytical models for use in pre-feasibility analysis and design of hydrokinetic schemes.
Principal Investigator: Dr Frederick Malan
Institution Name: University of Pretoria
Active Member Count: 8
Allocation Start: 2021-08-26
Allocation End: 2022-02-22
Used Hours: 51750
Project Name: Proton-Responsive Metal Complexes as Catalysts for Sustainable Hydrogen Production and Storage
Project Shortname: CHEM1358
Discipline Name: Chemistry
The research of the Malan Inorganic Chemistry Research group, led by Dr Frederick Malan (at the University of Pretoria), revolves around the design, economic synthesis, and characterization of new transition metal compounds. The catalytic activity associated with these compounds is evaluated mainly from a homogeneous catalysis point of view. Additional electrochemistry, computational chemistry, and single crystal X-ray diffraction studies aid in the understanding of the working of these catalysts. The aim of investigation of the role of proton-responsive ligands in CO2 and N2 fixation processes is to efficiently make use of these cheap building blocks as an entry point to fuel and other fine chemicals, as well as to curb the effect of greenhouse gas pollution. Catalysts reduce the activation energy by which a reaction takes place, and therefore the calculation and prediction thereof (via the CHPC) is important in this research. To date, promising results have been obtained with two PhD projects and several honours projects, with more results (especially from a computational point of view) anticipated for 2022 with the current students will be continuing this aspect of their research projects.
Principal Investigator: Dr Lawrence Borquaye
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 7
Allocation Start: 2021-08-27
Allocation End: 2022-02-23
Used Hours: 340604
Project Name: Biomolecular Computations
Project Shortname: HEAL1382
Discipline Name: Chemistry
This is the Borquaye Research Group in the Department of Chemistry, Kwame Nkrumah University of Science and Technology, Kumasi-Ghana. Our research group focuses on bioactive molecules such as natural products, peptides, essential oils and/or small molecules and their biological activities. We also make use of molecular docking and molecular dynamics tools to unravel potential modes of action of bioactive compounds, predict their preferred binding targets and explore events that mediate biomolecule-ligand interaction over a specified time period. The goal is to unearth compounds that could potentially be developed into drugs and to understand how these compounds work in biological systems. Our molecular docking and molecular dynamics projects require high performance computing, which we are able to access at the CHPC (Lengau). We have made significant progress and have submitted 2 manuscripts for publication based on data derived from the Lengau. We also have different projects in various levels of completion. We have been able to get a clear understanding of how certain natural products and peptides interact with their specific protein targets and biological membranes.
Principal Investigator: Prof Krishna Bisetty
Institution Name: Durban University of Technology
Active Member Count: 10
Allocation Start: 2021-08-27
Allocation End: 2022-02-23
Used Hours: 29100
Project Name: Computational Modelling & Bioanalytical Chemistry
Project Shortname: CHEM0820
Discipline Name: Chemistry
As an electrochemist and a computational scientist, our research work is mainly focused on smart-engineered nanostructured electrode materials, modified electrode surfaces and single particle characterization for the development of biosensors with applications in the environmental, food, health and pharmaceutical areas. The resources at CHPC are pivotal to establishing synergies between the experimental and computational approaches used to design electrochemical biosensors
Principal Investigator: Dr Fortunate Mokoena
Institution Name: North-West University
Active Member Count: 3
Allocation Start: 2021-08-30
Allocation End: 2022-02-26
Used Hours: 11161
Project Name: Protozoan parasites and cancer drug discovery
Project Shortname: CBBI1293
Discipline Name: Bioinformatics
The PI is Fortunate Mokoena, i am working on parasitic disease drug discovery projects requiring a far bit of molecular docking and dynamics simulations. We do not have enough computing units to run the calculations on either the local machines or institutional servers. We rely on CHPC to assist with the calculations. Virtual screening workflow, induced fit docking and molecular dynamics simulations. Most of the preliminary data has been generated.
Principal Investigator: Prof Joanna Dames
Institution Name: Rhodes University
Active Member Count: 7
Allocation Start: 2021-08-30
Allocation End: 2022-02-26
Used Hours: 71204
Project Name: Mycorrhizal Interactions
Project Shortname: CBBI0966
Discipline Name: Bioinformatics
The Mycorrhizal Research group is based at Rhodes University, Grahamstown.
Mycorrhizal fungi form a symbiotic relationship with the roots of the majority of plant species. The interaction between these soil fungi, other soil microbes and plants results in many beneficial growth effects making an important contribution to sustainable agriculture, horticulture, and environmental rehabilitation. There are several types of mycorrhizal relationships depending on the fungi and host plants involved. Little is known about the biodiversity of these fungi in South African soils. The use of next-generation sequencing provides an opportunity to unravel this biodiversity in order to better understand factors which impact on the relationship. The CHPC platform has assisted in the analysis of biodiversity of fungi in general and mycorrhizal fungi in particular as well as some interactions with soil bacteria.
Principal Investigator: Prof Zander Myburg
Institution Name: University of Pretoria
Active Member Count: 5
Allocation Start: 2021-08-30
Allocation End: 2022-02-26
Used Hours: 38027
Project Name: Forest Molecular Genetics (FMG) Programme
Project Shortname: CBBI1015
Discipline Name: Bioinformatics
The Forest Molecular Genetics (FMG) Programme at the University of Pretoria focuses on the genetic control of growth and development in fast-growing plantation trees with a view to enhance biomass production and improve wood properties for timber, pulp, paper, and biomaterials production. They work in close collaboration with South African forestry companies through the FMG Consortium (FMGC) to develop capacity and resources for the application of tree biotechnology in operational tree improvement programmes. The group has successfully used a systems genetics approach in an interspecific backcross population of E. grandis x E. urophylla to map key genomic regions affecting gene expression and metabolic profile associated with variation in growth and wood chemistry. They recently embarked on an effort to engineer cell wall traits such as xylan content and structure affecting pulp yield employing new approaches such as CRISPR-Cas9 genome editing. Over the past five years FMG have successfully used a single nucleotide polymorphism (SNP) marker chip with 60,000 DNA markers to genotype over 3000 Eucalyptus trees and recently also established a multi-species SNP genotyping chip for tropical pines. These SNP chips will aid in genome-assisted breeding projects in Eucalyptus and pine where the aim is to fast-track breeding cycles. The programme is also constructing a Genome Diversity Atlas for Eucalyptus and pine species grown in South Africa, laying the foundation for the emerging field of landscape genomics, which combines population genomics with analysis of interactions with environmental factors. Towards this, FMG is undertaking an effort to use long-range single-molecule sequencing to sequence genomes of several hundred Eucalyptus trees, also expanding to other genera including acacia, macadamia and pine. The programme has a strong collaborations with the Bioinformatics and Computational Biology Centre at the University of Pretoria where we access to a few servers and a cluster. We are also collaborating with Prof Jill Wegrzyn at the University of Connecticut, and have some access to the cluster at UConn. However, there is a need for additional resources, specifically for projects that rely on high-throughput processing of sequence data and applying machine learning.
Principal Investigator: Dr Emmanuel Sakala
Institution Name: Council for Geoscience
Active Member Count: 24
Allocation Start: 2021-08-30
Allocation End: 2022-03-14
Used Hours: 20920
Project Name: Application of high performance computing at the Council for Geoscience
Project Shortname: ERTH1227
Discipline Name: Earth Sciences
The Council for Geoscience (CGS) is mandated to develop and publish world-class geoscience knowledge products and to render geoscience-related services to the South African public and industry. For the CGS to fulfill this mandate and to advance the geoscience field within South Africa and beyond, the CGS is using the high-performing computer service to augment its computation capabilities. The sort of work that we are using the high-performing computing (HPC) resources is in the field of geophysics which involves collecting physical properties of the earth and using that information to guide the exploration of minerals and groundwater. The program is going well so far and we have managed to bring on board many of the young scientists to start using the HPC services.
Principal Investigator: Dr Zibo Keolopile
Institution Name: University of Botswana
Active Member Count: 3
Allocation Start: 2021-08-30
Allocation End: 2022-03-15
Used Hours: 204850
Project Name: Physics
Project Shortname: MATS1303
Discipline Name: Physics
Our research group is based in the Department of Physics, University of Botswana. It is focused on electronic structure calculations of molecular systems. Here we study gas phase molecular systems comprised a cluster of 1 to 10 molecules depending on the availability of research and computer manpower. Most calculations cannot be performed on normal workstation, but a supercomputer, therefore, a need for a more robust High Performance Computer. We have, so far, studied molecular trimers of neutral oxalic acid and the journal paper is underway. We will then focus our attention on anionic trimers of oxalic acid clusters.
Principal Investigator: Prof Graham Jackson
Institution Name: University of Cape Town
Active Member Count: 4
Allocation Start: 2021-08-31
Allocation End: 2022-02-27
Used Hours: 1149
Project Name: Insect neuropeptides
Project Shortname: CHEM1101
Discipline Name: Chemistry
Insect pesticides are non-specific and often harmful to beneficial insects like the honey bee, and humans. In this study we are aiming to develop species specific insecticides against the malaria mosquito, the desert locust and several fly pests.
Principal Investigator: Prof Phuti Ngoepe
Institution Name: University of Limpopo
Active Member Count: 9
Allocation Start: 2021-09-02
Allocation End: 2022-03-01
Used Hours: 6111317
Project Name: Computational Modelling of Materials: Mineral Processing
Project Shortname: MATS1404
Discipline Name: Material Science
The materials modelling centre at the University of Limpopo focuses on energy storage, alloys development and minerals modelling research. The mineral research deals with utilisation of computer software to investigate the properties of minerals and their surface reactivity. This extends to design of reagents used in the separation of minerals using flotation processes. The minerals modelling research has an impact in the minerals processing industry in the mines. The research works are compared with experiments through collaborations with other universities and companies, locally and international. In this light, experiments take times to get results and are costly. We utilizes the centre for high performance computing to do the simulations, which is cost effective. The minerals research projects are well progressing and a number of articles have been published and a number of students have graduated their masters and doctoral studies in this research field.
Principal Investigator: Dr Ouma Moro
Institution Name: North-West University
Active Member Count: 7
Allocation Start: 2021-09-03
Allocation End: 2022-05-05
Used Hours: 129337
Project Name: Rational design of novel catalysts using Chemistry 4.0 and 4IR tools
Project Shortname: MATS1365
Discipline Name: Material Science
Chemistry 4.0 is at the core of material informatics in the Fourth Industrial Revolution (4IR) space. Considering the COVID19 pandemic, high-performance computing has been valuable in solving real-life issues particularly when it comes to drug interactions in therapies being prescribed. This project has successfully investigated two cases of drug-drug interaction for COVID19 therapies. The project has also explored the use of machine learning to investigate oxide nanoparticles using quantitative-structure-activity relationships. At the core of this project is the application of density functional theory calculations to calculate descriptors that will be used as inputs to machine learning models and also for training the machine learning (ML) and deep learning (DL) models. Those working on the project include Dr Cecil NM Ouma (the Principal Investigator with vast experience in ab initio modelling), Gladys Kingóri, Umer Ahmed and Makola Lekgowa (PhD students) and Lele Ntando and Malati Majahekupheleni (MSc students). The HPC resources from CHPC in Cape Town have been at the core of all the research that we have done so far. It is these resources that are both facilitating and mediating Africa's concerted efforts toward catching up with developed nations in terms of research, development and capacity building.
Principal Investigator: Dr Aniekan Ukpong
Institution Name: University of KwaZulu-Natal
Active Member Count: 0
Allocation Start: 2021-09-07
Allocation End: 2022-03-15
Used Hours: 435384
Project Name: Theoretical and Computational Condensed Matter and Materials Physics
Project Shortname: MATS0941
Discipline Name: Physics
I am Dr. Aniekan Magnus Ukpong, and I am the lead researcher in the Theoretical and Computational Condensed Matter and Materials Physics Research Group, as well as the Principal Investigator in the CHPC's Research Program: MATS0941. We are now undertaking the computational study of Mott insulators, a class of materials that are expected to conduct electricity according to the band theory of solids but turn out to be insulators at low temperatures. This is because materials with strong magnetic interactions and antiferromagnetic ground states in whose proximity other exotic phases can emerge are attracting current interest for fundamental scientific reasons and because they have applications as emergent materials in thin-film magnetic heterostructures that show potential in high-temperature superconductivity. Mott insulators are not correctly described by band theory of solids due to their strong electron-electron interactions, and they exhibit a Mott transition from the metal to insulator state due to the strength of electronic interactions. This can create a distinct magnetic order at sufficiently low temperatures, but an exception is a quantum spin liquid (QSL), which is characterized by the presence of no spontaneous symmetry breaking, and magnetic quasiparticles, called spinons. The fascinating and fertile situation of the QSL ground state occurs in Mot insulators when spin frustration is present. Strong quantum fluctuations are enhanced by its intrinsic spin frustration to hinder a conventional magnetic ordering, and this reduces the magnetic transition temperature to 0 K. The ingredients of Mott Physics occur in the κ-phase of κ-(BEDT-TTF)2Cu2(CN)3, where, BEDT-TTF stands for bis-ethylenedithio-tetrathiafulvalene, hence we study it as a model material. Our main modeling tool is still quantum mechanics as formulated within density functional theory, and we implement it computationally using the HPC platform. Calculations of the ground state total energy at zero temperature are now modified to include the effects of finite temperature and pressure, and this framework allows us to determine the partition function. This allows for important thermodynamic properties of our model material to be obtained from HPC calculations, as a research alternative to using expensive instruments for experimentation. Our findings could pave the way for the systematic development of a new family of paramagnetic salts that are suitable for achieving high-temperature superconductivity, in a cost-effective manner, exclusively through HPC.
Principal Investigator: Prof Koop Lammertsma
Institution Name: University of Johannesburg
Active Member Count: 3
Allocation Start: 2021-09-07
Allocation End: 2022-03-06
Used Hours: 303823
Project Name: ChiralCat
Project Shortname: CHEM1410
Discipline Name: Chemistry
The ChiralCat project is conducted at the University of Johannesburg under the direction of Prof. K. Lammertsma with Prof. A. Muller as co-supervisor and Dr. G. Dhimba as computational chemist and with PhD and MSc students performing experimental studies.
ChiralCat is about chiral-at-metal catalysis in which the chiral integrity of the transition metal catalysts is maintained during a chemical reaction. The rational design of such catalysts capable of effecting enantioselective chemical transformations is of paramount importance to satisfy the increasing industrial demand for chiral fine chemicals. ChiralCat is an innovative approach in asymmetric catalysis, requiring a detailed understanding to advance the field. So far, asymmetric catalysis is dominated by catalysts carrying expensive chiral ligands. Such conventional catalysts require exhaustive screening of the chiral ligand pool to obtain products with high enantiomeric excess. Not only is this a tedious and costly process, also the chiral ligands are often far more expensive than the transition metals. ChiralCat explores instead the use of abundantly available transition metals with readily available simple ligands to compose catalysts that are inherently chiral and that keep their chiral integrity during the catalytic reaction.
To provide these insights and assist experimentalist in synthesizing chiral-at-metal catalysts requires insight in the molecular behavior of the catalysts and their catalytic reactions. Computational chemistry is by far the best and most effective means to provide this insight, which may well simplify many industrial processes. Because of the available and indispensable compute power of CHPC we could make much progress in the first six months of this project showing the feasibility of asymmetric epoxidation of olefins with a simple molybdenum catalyst.
Principal Investigator: Dr Mehdi Mehrabi
Institution Name: University of Pretoria
Active Member Count: 1
Allocation Start: 2021-09-03
Allocation End: 2022-03-02
Used Hours: 60284
Project Name: Modeling of heat transfer characteristics and pressure drop of nanofluids in micro and mini tubes.
Project Shortname: MECH1086
Discipline Name: Other
We in the clean energy research group (CERG) at the University of Pretoria are working on the new generation of heat transfer fluid to increase the efficiency of the cooling systems. Nanofluids are the new generation of fluids that can give us higher thermal conductivity compared to that of the conventional fluids. But the issue is that they also increase the pressure drop and consequently the required pumping power. Their thermophysical proprieties are also are not as straight-forward as the conventional fluid to predict. The current project is trying to close this gap by providing the better understanding of the heat transfer characteristics and pressure drop of the nanofluids.
Principal Investigator: Dr ADEDAPO ADEYINKA
Institution Name: University of Johannesburg
Active Member Count: 7
Allocation Start: 2021-09-09
Allocation End: 2022-03-08
Used Hours: 342262
Project Name: Computational Design of Molecules
Project Shortname: CHEM1221
Discipline Name: Chemistry
The Computational Design of Molecules research group is based at the University of Johannesburg. We aim to design new and improved molecules and materials for applications in Catalysis and renewable energy. Since the availability of energy is one of the main challenges of the African continent, being able to achieve our aims as a group will provide clean energy solutions for the continent. We use computational chemistry software to explore the properties of molecules which is responsible for their activity and then use the knowledge gained to design more efficient and improved materials.
Principal Investigator: Prof Obodo Kingsley
Institution Name: North-West University
Active Member Count: 14
Allocation Start: 2021-09-09
Allocation End: 2022-03-08
Used Hours: 2055363
Project Name: Ab initio modelling of liquid organic hydrogen carriers catalyts and 2D materials
Project Shortname: MATS1366
Discipline Name: Physics
The current research programme "MATS1366: Ab initio modelling of liquid organic hydrogen carriers catalyst and 2D materials" was set up to carry out research activities in the area of catalyst development and other two-dimensional materials. The principal investigator is based in HySA infrastructure center of competence in the North West University, South Africa. To date, the members of the research program include: Kingsley Obodo, Hailouf Houssam, Ijeoma Onyia, Stella Ogochukwu, etc. The use of the CHPC facility is enabling the members to undertake in their Masters as well as Ph.D research work. The group members are undergoing significant progress in their work as demonstrated by the novel research generated, which is currently under-review and submitted for publication. Some of the work published during the reporting period are Sn3C2 monolayer with Transition Metal Adatom as well as the dopant complexes and the pristine Ti3N2 MXene monolayer using spin-polarized density functional theory calculations with van der Waals correction. Other work under review and to be submitted include but not limited to the catalyst for liquid organic hydrogen carriers (LOHCs) de-hydrogenation as well as catalysts for water electrolysis. Water electrolysis entails the breaking down of water to give hydrogen and oxygen. The main technique applied is density functional theory as implemented in CASTEP, Quantum Espresso, GPAW and VASP packages.
Principal Investigator: Prof Arnaud Malan
Institution Name: University of Cape Town
Active Member Count: 7
Allocation Start: 2021-09-09
Allocation End: 2022-03-08
Used Hours: 309799
Project Name: UCT CFD
Project Shortname: MECH1314
Discipline Name: Computational Mechanics
InCFD used CHPC to do several simulations of sloshing tanks to compute the forces and slosh damping using their in-house software Elemental. Elemental's excepational capability was proven via simulating two violent slosh experiments viz. lateral and verticle slosh in the presence of gravity. Particularly high accuracy was demonstrated, where the L2 error norm in the latter case is less than 2%. It is further shown that accounting for gas compressibility in a weak sense reduces predicted pressure peaks by up to 50% while the solver is 7 times faster to run. This is seen as a key development for more effective slosh impact design.
Principal Investigator: Mr Anban Pillay
Institution Name: University of KwaZulu-Natal
Active Member Count: 8
Allocation Start: 2021-09-10
Allocation End: 2022-03-09
Used Hours: 22099
Project Name: Deep Learning Approaches
Project Shortname: CSCI1078
Discipline Name: Computer Science
The Deep Learning research group belongs to the Centre for Artificial Intelligence Research Node at the University of KwaZulu-Natal. The Centre for Artificial Intelligence Research (CAIR) (CAIR.za.net) is a South African distributed Research Network that conducts foundational, directed and applied research into various aspects of Artificial Intelligence. The research group conducts foundational research in machine learning in general and deep learning in particular. Deep learning projects require substantial computing resources and the CHPC is vital to its success. The projects involve various area of deep learning that include deep reinforcement learning, convolutional and recurrent neural networks. Application areas include image processing, natural language processing and speech recognition. Several students are busy with projects at present.
Principal Investigator: Dr Abdulrafiu Raji
Institution Name: University of South Africa
Active Member Count: 3
Allocation Start: 2021-09-13
Allocation End: 2022-03-12
Used Hours: 286487
Project Name: Magnetocaloric effect in selected metallic nanoclusters on two-dimensional (2D) substrates, and in selected rare-earths.
Project Shortname: MATS1156
Discipline Name: Material Science
The research study is based at the Department of Physics of the University of South Africa (UNISA) and is being led by Dr. Abdulrafiu Raji. The study focus is on computational studies of two-dimensional (2D) crystals, nanocrystal and organic semiconductors for potential applications in spintronics and catalysis. Other applications being envisaged are ultrathin refrigeration and air-conditioning systems, and any nanoarchitecture where cooling may be needed. Specifically, the research aims to induce novel electronic and magnetic interactions in nanostructured materials which could be exploited for practical applications. For example magnetocaloric effect is a property of magnetic materials which could be exploited for refrigeration purposes. While the conventional refrigerator system is based on compression and evaporation of often environmentally unfriendly gases, magnetic refrigeration systems is based on magnetizing and demagnetizing a magnetic material. In the last few years, the focus in magnetocaloric research seems to have shifted toward ultrathin materials. Because of plethora of candidate materials that may be investigated as possible candidates for MCE, performing experimental investigations may be time and resources consuming. Therefore, computational studies could provide the lead and narrow the range of materials that could be considered for further experimental investigations. Performing ab-initio density-functional theory (DFT) study to determine the properties of candidate material systems is a viable and state-of-the-art approach which can complement experimental efforts. DFT calculations require specialized softwares as well as high-capacity data storage and memory requirements which are often beyond the capability of an ordinary desktop computer. The softwares have been made available on the CHPC platform. The CHPC administrators also maintain the softwares and provide different kinds of technical supports to students involved in the projects. Thus, the availability of high-performance computing (HPC) facility is absolutely necessary. The project currently have postgraduate students working on its several aspects.
Principal Investigator: Prof Yin-Zhe Ma
Institution Name: Stellenbosch University
Active Member Count: 4
Allocation Start: 2021-09-14
Allocation End: 2022-03-13
Used Hours: 136591
Project Name: Machine learning in 21-cm cosmology
Project Shortname: ASTR1323
Discipline Name: Astrophysics
I am simulating the epoch of reionization to study the effect of mini-halos (cooling via molecular ways instead of atomic cooling ways in normal halos) in the early universe. Then I use multinest sampling to study the Bayes evidence of the model including mini-halos. Meanwhile, I also calculate the CMB power spectra of E-mode and temperature and E-mode correlation
Principal Investigator: Dr Gavin Gouws
Institution Name: SAIAB
Active Member Count: 9
Allocation Start: 2021-09-15
Allocation End: 2022-03-14
Used Hours: 55482
Project Name: RADseq studies of aquatic biodiversity
Project Shortname: CBBI1274
Discipline Name: Other
The Marine Fish Research Team in the Aquatic Genomics Research Platform at the National Research Foundation - South African Institute for Aquatic Biodiversity employs a variety of genetic tools to understand the evolution and ecology of marine fishes (and marine invertebrates). We are interested in the relationships between species, populations and individuals, and in the movement and migration of individuals, with a view to informing conservation and management. Over the last year, we have conducted studies examining whether the genetic diversity of commercially-exploited marine fishes have changed over time, as a result of harvesting, and whether an endangered line-fish exists as and should be managed as a single population along the South African coast. For the former, SNP genotype data were generated and analysed using applications on the CHPC, while the latter used the CHPC and it's applications to model patterns of migration, dispersal and changes in population size using mitochondrial and nuclear genetic data. Some of this research has recently been published, with additional manuscripts currently being prepared.
Principal Investigator: Prof Andries Engelbrecht
Institution Name: Stellenbosch University
Active Member Count: 13
Allocation Start: 2021-09-15
Allocation End: 2022-06-23
Used Hours: 1764245
Project Name: Computational Intelligence for Optimization
Project Shortname: CSCI0886
Discipline Name: Computer Science
Algorithmic models of social organisms in nature, such as bird flocks, have been developed to solve complex optimization problems. Real-world problems include knowledge discovery from datasets, data clustering and solving portfolio optimization problems and other single-objective optimization problems. Before these algorithms can be used to solve real-world optimization problems, it is necessary to gain a clear understanding of how they work, and to gauge their performance on benchmark problems, in comparison with established algorithms. The empirical analyses required is computationally expensive, and take excessively long on standard desk top computers. The CHPC provides a means to reduce the time to quantify the performance of these algorithms prior to deployment to solve real-world problems.
Principal Investigator: Mr Joachim Huyssen
Institution Name: University of Pretoria
Active Member Count: 2
Allocation Start: 2021-09-17
Allocation End: 2022-10-31
Used Hours: 6173
Project Name: Aircraft Configuration Investigations
Project Shortname: MECH1372
Discipline Name: Computational Mechanics
Computing cluster helps compare a new airline body against the old.
Since the introduction of the Boeing 707 around 1960, the layout of the airliner has remained essentially unchanged. Yet, the question, of what an aircraft should look like to be most efficient has not found consensus in the aeronautical community. Some radical proposals for alternative aircraft configurations have emerged over the last four decades but none of these has changed the industry. Given the adverse impact on global climate by this sector of the transport industry, the notion of sustainable aviation keeps research on alternatives ongoing.
The aeronautics research in the Department of Mechanical and Aeronautical Engineering at the University of Pretoria participates in this exploration of alternative aircraft configurations. This work has led to an international patent which describes an aircraft arrangement that is not radically different from the current typical aircraft. As such, it can be used in any field of aviation. The most prominent difference shows a fuselage that is shorter and wider than its traditional counterpart. This fuselage makes an important contribution to the lifting force of the aircraft and its stability while being minimized for mass and air resistance. Furthermore, the future airliner may fly without the large vertical stabilizer. In consequence, it may fly the same mission with substantially less fuel. A master's student from the University of Pretoria is using the power of the computing cluster to run large simulations to compare the predicted air resistance of the proposed fuselage to estimates of the typical airliner fuselage. He is currently running benchmark studies for which reliable values are available for comparison to his simulations. He will then simulate and refine the reshaped fuselage which would carry the same payload as its counterpart. Performance assessments for the 707 had to come from actual tests. Today reliable predictions can be computed long before committing to prototypes.
This requires a virtual aircraft to be surrounded by millions of virtual cells for each of which a supercomputer needs to predict the pressure and velocity. Thereby, the forces on the aircraft can be estimated over its entire surface without impairing the cost and risk of building and flying a new design.
Principal Investigator: Dr Anna Bosman
Institution Name: University of Pretoria
Active Member Count: 13
Allocation Start: 2021-09-17
Allocation End: 2022-03-16
Used Hours: 561586
Project Name: Deep Learning and Neural Network Research at CS UP
Project Shortname: CSCI1166
Discipline Name: Computer Science
The research program of the Computer Science Department, University of Pretoria (UP), is headed by Dr Anna Bosman, and forms a part of the Computational Intelligence Research Group (CIRG) at UP. The focus of the program is on the following research topics: fitness landscape analysis of neural networks and real-life applications of deep learning. The following deep learning applications are currently investigated: (1) convolutional neural networks for satellite image analysis; (2) evolution of recurrent network architectures; (3) convolutional neural networks for image segmentation; (4) convolutional neural networks applied in the entomology domain. Studying fitness landscapes of neural networks deepens our general understanding of the fundamental principles of neural network research, and enables progress in fundamental artificial intelligence. Deep learning applications allow us to apply modern artificial intelligence techniques in the South African context. Thus, the usage of CHPC helps us strengthen machine learning and artificial intelligence research in South Africa. Sampling and estimation techniques for fitness landscape analysis are used to probe the error surfaces of the deep and shallow neural networks. The research is of empirical nature, and entails heavy sampling of the search space. Such experiments would not be feasible without access to a computer cluster. Deep learning applications benefit significantly from being run on a GPU, therefore, access to the GPUs allows us to upscale the experiments to real-life applications. The research program yields annual publications in highly ranked international journals and conferences. This indicates that the research program is productive, and successfully delivers academic outputs.
Principal Investigator: Dr Gerhard Venter
Institution Name: University of Cape Town
Active Member Count: 10
Allocation Start: 2021-09-17
Allocation End: 2022-03-16
Used Hours: 833346
Project Name: Simulation of Ionic Liquids
Project Shortname: CHEM0791
Discipline Name: Chemistry
Dr Venter's research group is part of the Scientific Computing Research Unit at the University of Cape Town that uses computer simulation to study the properties of a special class of solvents, called "ionic liquids". A fundamental understanding of the physical properties of ionic liquids and transports properties of solutes in these systems can lead to the rational design of new, environmentally friendly liquids with applications as electrolytes in next-generation batteries and energetic materials. The simulations not only provide first-principles characterization of ionic liquids and solutes, but also form the basis of models that aim to use machine learning for the prediction of thermodynamic properties.
High performance computing is required for all research conducted in this group and without the resources and infrastructure provided by the Centre for High Performance Computing (CHPC), the research will not be possible.
Principal Investigator: Dr Henry Otunga
Institution Name: Maseno University, Kisumu, Kenya
Active Member Count: 4
Allocation Start: 2021-09-20
Allocation End: 2022-04-21
Used Hours: 72845
Project Name: Computational Study of Hematite Clusters
Project Shortname: MATS0888
Discipline Name: Material Science
This is the Computational Materials Science Group of Maseno University, Department of Physics and Materials Science. I am the lead researcher in this group. Other members of the group are Nicholas Ong'wen (a PhD student, Maseno University), Richard Onyango (an MSc student, Maseno University) and Benjamin Omubandia (also an MSc student, Maseno University).
Sunlight is the most abundant renewable energy resource and considered to be the ultimate solution to address the global energy problem: However, the vision of solar energy providing a substantial fraction of global energy infrastructure is still far from being realized. The major challenge is to develop an efficient and cost-effective approach for storing solar energy that can be used on demand on a global scale. Splitting water photoelectrochemically is a promising way in which solar energy can be harvested and stored through photocurrent-driven chemical bond rearrangement leading to the formation of molecular hydrogen (2H2O→2H2+O2, E0=1.23 eV, Standard Hydrogen Electrode, SHE), which can be used directly as fuel or as an intermediate in the production of other fuels.
Hematite (alpha-Fe2O3) has emerged as a candidate photoanode material due to several favorable features: a band gap of about 2.0 eV (which allows for a large fraction of solar spectrum to be absorbed), excellent stability under aqueous environments, and abundance in nature. Converting energy from the Sun to electricity efficiently via photoelectrochemical splitting of water is based on photo-induced charge separation at an interface (semiconductor/water interface) and the mobility of photogenerated carriers (electron-hole pairs). However, in bulk hematite, the separated charges recombine very fast (very short lifetimes of electron-hole pairs) and have very low mobility, hindering the conversion efficiency. These limitations can be dealt with by nano-structuring hematite, by growing ultrathin films of hematite, and by introducing impurity atoms in hematite. In this work, the structure and energetics of various size clusters of Fe2O3 have been investigated using the plane-wave pseudopotential approach to Density Functional Theory (DFT) and ab initio Molecular Dynamics (MD).
These calculations require a significant amount of compute resources in terms of memory and cpu scaling. In this respect, I am grateful to the CHPC for granting me access to their supercomputing facilities and I hope for more collaborations in the future.
Principal Investigator: Mr Mogesh Naidoo
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 7
Allocation Start: 2021-09-20
Allocation End: 2022-03-19
Used Hours: 34866
Project Name: Simulating atmospheric composition
Project Shortname: ERTH0846
Discipline Name: Earth Sciences
Air pollution can have large negative impacts on human health, agriculture, ecosystems, visibility and climate. In South Africa, although ambient air quality is regulated, many areas are still out of compliance with the National Ambient Air Quality Standards. In order to protect human health and mitigate impacts, it is critical to improve air quality. The Constitution provides that everyone has a right to have an environment that is not harmful to their health. The Atmospheric Composition Focus Area (ACFA) in the CSIR Climate and Air Quality Modelling research group aims to provide the evidence base to quantify the impacts of air quality and to improve air quality.
The group uses the CHPC to run an air quality model to simulate urban and regional air quality at high resolution. This chemical transport model simulates the physics and chemistry of the atmosphere. The processes represented within the model are complex, and thus computationally intensive, which makes use of the CHPC facility a necessity. The CSIR ACFA is the only group in South Africa routinely running a chemical transport model to simulate air quality for management purposes.
Using the CHPC resources, the team has been able to simulate the impact of policy interventions on air quality in cities. Additionally, the team has simulated the health risk from air pollution regionally in South Africa. In the past, this has been done using monitoring station data only, which then limits the analysis to only those living directly around the station. The team has also simulated the impact of climate change on air pollution. These simulations use the climate projections from the CSIR's CCAM-CABLE, which is also run at CHPC, to provide meteorology input into the air quality model.
These outputs directly provide the evidence base needed for decision makers to draft and implement policies and interventions to effectively improve air quality as well as understand its impacts, now and into the future.
Principal Investigator: Dr Geoff Nitschke
Institution Name: University of Cape Town
Active Member Count: 14
Allocation Start: 2021-09-21
Allocation End: 2022-03-20
Used Hours: 1266012
Project Name: Evolving Complexity
Project Shortname: CSCI1142
Discipline Name: Computer Science
Experiments aim to use large, high volume real-world data sets, such as those from the medical industry (e.g. patient MRI scans), to supervise the training of machine learning methods that can automatically formulate a prognosis or offer a complementary hypothesis to medical experts. Besides the medical domain, new machine learning methods are being tested on similarly large data-sets in other domains including: robotic control, autonomous vehicles, and cybersecurity systems.
Principal Investigator: Dr Quinn Reynolds
Institution Name: Mintek
Active Member Count: 6
Allocation Start: 2021-09-23
Allocation End: 2022-04-08
Used Hours: 256724
Project Name: Computational Modelling of Furnace Phenomena
Project Shortname: MINTEK776
Discipline Name: Applied and Computational Mathematics
The world of pyrometallurgy, the subtle art and science of using high temperatures to produce the metals and other mineral products that shape our industrialised lives, presents a number of unique challenges to engineers. Imagine having to design, build, and operate a complex piece of machinery without ever being able to get close or see inside it while it is working! But at Mintek that is exactly what we try to do for metallurgical smelting furnaces - their interior conditions are so harsh and hazardous to both people and instrumentation that we can often rely only on our fundamental knowledge of the various processes and phenomena at work to understand them properly. A key tool in building that understanding is computational modelling - harnessing the power of massive computers like those at CHPC to solve the enormous interconnected sets of equations that define how physics and chemistry work under the intense conditions inside furnaces. And once that deep understanding is in place we can start the exciting work of engineering the newer, cleaner, greener, and more efficient pyrometallurgical technologies of the future.
Principal Investigator: Prof Cedric McCleland
Institution Name: Nelson Mandela Metropolitan University
Active Member Count: 1
Allocation Start: 2021-09-21
Allocation End: 2022-04-28
Used Hours: 8406
Project Name: Reactive Intermediates
Project Shortname: CHEM1394
Discipline Name: Chemistry
This project entails theoretical studies on certain reactive intermediates that play key roles in many chemical reactions. Such reactions are important for the synthesis of numerous compounds that find wide application in organic chemistry, including as pharmaceuticals. Our focus has been on alkoxyl radicals and radical cations derived from unsaturated alkanols. It is crucial that the characteristics of these intermediates, as well as the mechanisms by which they react, be thoroughly understood to ensure their optimal use. To this end, we have undertaken an extensive quantum mechanical study at the density functional (DFT) level of theory on the title systems. Various functionals (B3LYP and ωB97X-D, the latter also accounting for dispersive interactions) and basis sets ranging from 6-31G* to larger variants incorporating diffuse and polarization functions, have been applied in the study. The results provide detailed insight into the factors that influence cyclisation regioselectivities and in particular the mechanisms by which radical cations undergo nucleophilic ring closure. Modelling of substituent effects in ethylene and benzene radical cations have also been carried out, providing understanding of how a wide range of substituents influence reactivity in these systems. The study is ongoing and has already led to the successful completion of a PhD thesis.
Group members: Prof CW McCleland and Dr PMC Lee, Chemistry Department, Nelson Mandela University, Port Elizabeth.
Principal Investigator: Dr Jennifer Veitch
Institution Name: SAEON
Active Member Count: 5
Allocation Start: 2021-09-23
Allocation End: 2022-03-22
Used Hours: 57677
Project Name: SAEON Coastal and Regional Ocean Modelling Programme
Project Shortname: ERTH1103
Discipline Name: Earth Sciences
Coastal and regional oceans are often poorly represented in current global ocean models. This is usually due to use of coarse resolutions in the models subsequently leading to shortcomings in representation of fine scale coastal and regional processes. To improve this regional, high resolution or nested configurations are commonly applied to capture and study these processes. In this programme, high resolution ocean models (NEMO and ROMS) are used to understand physical and biogeochemical processes around the coast of South Africa and adjacent basins. These models contain a suite of useful schemes that need to be tested and used in a regional context. Once configured and tested these models can be used to improve knowledge by taking advantage of available in-situ and satellite data to understand oceanic processes under realistic scenarios.
Principal Investigator: Prof Sybrand van der Spuy
Institution Name: Stellenbosch University
Active Member Count: 5
Allocation Start: 2021-09-22
Allocation End: 2022-03-21
Used Hours: 160154
Project Name: CFD simulation of turbomachinery
Project Shortname: MECH1187
Discipline Name: Computational Mechanics
The turbomachinery research group at Stellenbosch University has been making use of the CHPC for their CFD simulation work during the past three years. The use of the CHPC enables the accurate and detailed modelling of aspects like gas turbine combustion and fan noise. The results achieved to date have been world class.
Principal Investigator: Prof Kevin Naidoo
Institution Name: University of Cape Town
Active Member Count: 10
Allocation Start: 2021-09-24
Allocation End: 2022-03-23
Used Hours: 813982
Project Name: Reaction Dynamics of Complex Systems
Project Shortname: CHEM0840
Discipline Name: Chemistry
Professor Kevin Naidoo has held the SARChI Scientific Computing research chair since the award in 2007. The chair was jointly established in a collaboration with UCT and the CHPC. Professor Naidoo's group develops cutting edge chemical, chemical biological and biological software packages that enables the discovery of Drugs and Diagnostics. Following the core development in the Scientific Computing Research Unit (SCRU) at UCT his group tests the packages and then applies the computer models using the CHPC's GPU and CPU clusters to urgent healthcare problems focused on cancer and respiratory diseases SCRU develops therapeutics and diagnostics for cancer and respiratory diseases. This has lead to Professor Naidoo being the PI on a Phase 1 breast cancer diagnostic trial where the groups bioinformatics and machine learning methods have led to a potential breast cancer biomarker. Further the development of cancer and respiratory therapeutics are being experimentally tested. The combined high development of SCRU software and the power of the CHPC's compute environment has enabled Professor Naidoo to establish a world first translation compute to clinic platform.
Principal Investigator: Mr Adebayo Azeez Adeniyi
Institution Name: University of KwaZulu-Natal
Active Member Count: 7
Allocation Start: 2021-09-27
Allocation End: 2022-03-26
Used Hours: 150049
Project Name: Drug Discovery Research Using Quantum and Molecular Dynamic Simulation
Project Shortname: CHEM0958
Discipline Name: Chemistry
CHPC is the major facility that makes the progress we have made in our research work feasible. Through the CHPC, we have made a significant research impact and have enhanced our productivity. Just this year alone, we have recorded 2 published articles. With the available facilities from CHPC, we have been able to expand our research focus to include vaccine development, photochemistry and polymer chemistry in addition to our area of research on drug development and electrochemical properties of small molecules. We made use of several packages that are installed on CHPC for Bioinformatics, Quantum, and Molecular Dynamic Simulation. The reason for expanding our research focus is based on the pressing need of the society and global challenges, that lead us into epitopes-based vaccine development from the genomes of the pathogens, polymer chemistry, photochemistry of small molecules for application as photoactive compounds, and molecular electrochemistry. Our research uses theoretical modeling to give rational insight into identified problems which can help in the experimental designs for possible solutions. CHPC is an important facility to our research group, we make use of the facilities and the software like Gaussian, Orca, Games and Newchem; also used molecular dynamic packages like Gromacs, Amber and Lammps. We have applied quantum packages to study the chemical and spectroscopic properties of small molecules of which we have been able to reproduce some of the experimental results.
It is obvious that our research progress and achievement revolve around the service of CHPC and much of our research output would not have been possible without the support from CHPC.
Principal Investigator: Dr Adam Skelton
Institution Name: University of KwaZulu-Natal
Active Member Count: 3
Allocation Start: 2021-09-22
Allocation End: 2022-03-21
Used Hours: 5255
Project Name: Molecular modelling of biomolecules and materials
Project Shortname: CHEM0798
Discipline Name: Chemistry
I am Dr. Adam Skelton and I work at the College of Health Science university of KwaZulu-Natal. I work on understanding the behaviour of biological and material systems on the molecular level.
A particular interest is in using molecular simulation to direct the rationale design of functionalized nanoparticles for a range of applications such as drug delivery, water treatment and catalysis. In particular, drug delivery is the use of nanoparticles to carry antiviral or antibacterial drugs into parts of the body that are difficult for conventional drugs to reach. Drug-delivery is, therefore, of huge importance in South Africa, especially in the battle against HIV and Tubercolosis.
The research requires running a series of computar simulations, which may take hours at a time. CHPC has had a great impact since it provides me with the facilities to perform such simulations.
Principal Investigator: Prof Phuti Ngoepe
Institution Name: University of Limpopo
Active Member Count: 21
Allocation Start: 2021-09-27
Allocation End: 2022-06-02
Used Hours: 1972060
Project Name: Computational Modelling of Materials: Energy Storage
Project Shortname: MATS0856
Discipline Name: Material Science
The Research Group is based in Materials Modelling Centre at the University of Limpopo, under the leadership of Professor Phuti Ngoepe. Energy Storage Systems: We have, through simulations on primary particles of lithium-manganese-oxide (LMO) spinel, characterized disruptive transition during discharging; and are currently exploring porosities to enhance capacity of such cathodes. All nickel manganese oxide (NMC) primary nanoparticles have been modelled modified to NMC and the nature of structural disorder associated with discharging are studied. This will inform the new manganese rich NMC co-precipitation experiments set up on the UL campus. In addition, stable O3-type structures with enhanced performance, are predicted by doping using 4IR machine learning methods. Approaches of minimizing anion oxidation at surfaces during discharging are explored. Some aspects of the beyond lithium ion batteries are continuing in the form of Li-S,Se, particularly on the S rich side, together with work on catalysts for Li,Na,K,-air batteries.
Metal-air batteries are attractive for any application where weight is a primary concern, such as in mobile devices. Since oxygen doesn't need to be stored in the battery, the cathode is much lighter than that of a lithium-ion battery, which gives lithium-air batteries their high energy density. Most of the previous studies were limited by applying pure oxygen in the cathode, without putting into consideration the effect of the catalyst, which will play a significant role in Oxygen Reduction Reaction (ORR) and Oxygen Evolution Reaction (OER). Adsorption of oxygen at (110) Na-MO2 and Li-MnO2 was investigated, using density functional theory (DFT) calculations, which is important in the discharging and charging of both Li-air and Na–air batteries. Oxygen adsorption on Li/MO2 was simulated and it was found that in all the metal oxides (MnO2, TiO2 and VO2) comprises most stable orientation is the dissociated composition where there is an oxygen atom on the "bulk-like" positions on top of each of the M cations. Adsorption of oxygen on Na/MO2 was also investigated and it was observed that the catalysts encourage formation of the discharge product reported in literature, i.e. NaO2. The surface NaO2 appeared to have comparable bond lengths reported for monomer NaO2.
The computational modelling studies, through CHPC, are intended to predict processes that are essential in pilot and production plants for advanced battery systems, mineral processing and metal production industries. Thus far the progress on such studies is excellent and CHPC HPC clusters have been utilised to extensive multiscale studies directly and through external interfaces such as Materials Studio and MeDeA. All in all, the developments contributes extensively to human capacity/capital development.
Principal Investigator: Prof Vinesh Maharaj
Institution Name: University of Pretoria
Active Member Count: 2
Allocation Start: 2021-09-28
Allocation End: 2022-03-27
Used Hours: 20423
Project Name: Biodiscovery programme
Project Shortname: CHEM1458
Discipline Name: Chemistry
The Biodiscovery research group at the University of Pretoria, Department of Chemistry, conducts research on natural product compounds against a variety of different diseases including COVID19, cancer, Malaria and various other diseases. As part of the research conducted, the research team makes use of the CHPC to perform in-silico docking work to determine and substantiate any observed activity in in-vitro assays. MD simulations, being the most computationally exhaustive, are performed using the CHPC's resources, as we do not have these. The MD simulations are greatly required for any publications. Most ligand docking is performed on the researchers own PCs, using the Schrodinger license, however MD simulations are performed on the CHPC to enable quick results. Most journals require some sort of MD work done to substantiate any observed activity against the target. The project is progressing well with some publications expected to be finalized in the coming year.
Principal Investigator: Prof Zenixole Tshentu
Institution Name: Nelson Mandela Metropolitan University
Active Member Count: 9
Allocation Start: 2021-10-04
Allocation End: 2022-04-02
Used Hours: 96369
Project Name: Metal-ligand, anion-cation interactions and protein-ligand studies
Project Shortname: CHEM0864
Discipline Name: Chemistry
The CHPC resources have assisted us significantly at the NMU Chemistry Department to implement our research programme on design of reagents for metal ion or complex anion specificity. The outer sphere coordination has been studies computationally through the use of non-covalent interactions and electrostatic potentials. This work has focused on cationic interaction with IrCl62- or PtCl62- to study factors that lead to selective interactions. The results show promise for the design of such reagents and this is a great contribution towards the hydrometallurgical processing of precious metals.
Principal Investigator: Dr Lucy Kiruri
Institution Name: Kenyatta University, Nairobi, Kenya
Active Member Count: 5
Allocation Start: 2021-10-05
Allocation End: 2022-04-03
Used Hours: 105302
Project Name: KU Computational Chemistry Research Group
Project Shortname: CHEM1032
Discipline Name: Chemistry
The members of the group are in two different areas of Molecular modeling, that is; quantum mechanical calculations and Molecular Dynamics. For the QM calculations, we largely use the Gaussian 16 program and for MD simulations we use the GROMACS 2020. We research drug discovery for infectious and non-infectious diseases like Parkinson's disease. Our research on Parkinson's is motivated by the discovery of small molecules (natural products) whose activity was observed to be high in vitro and in vivo, however, the actual model and mechanisms of actions remain not well explored. We have employed a computational approach to understanding the mechanisms of action. Although our preliminary results are promising and help to explore the underlying mechanisms, we are still exploring many other mechanisms of action of the drug-like molecules.
Principal Investigator: Dr Matshawandile Tukulula
Institution Name: University of KwaZulu-Natal
Active Member Count: 1
Allocation Start: 2021-10-05
Allocation End: 2022-04-03
Used Hours: 4286
Project Name: Medicinal Chemistry and Computer-Aided Drug
Project Shortname: HEAL1346
Discipline Name: Chemistry
One of the main topics we working in our research group incorporated the development of new antimalarials. This involves looking at a new target that is recently been identified as key target in malarial proliferation (i.e., Plasmodium falciparum phosphatidylinositol 4-kinase (PfPI4K)). PfP14K is a lipid kinase which regulates intracellular parasite activities in all stages of the parasite growth and development of the Malaria parasite within a human host. Currently, there are only two related classes of inhibitors namely (i) the imidazopyridine/pyrazines/pyradazines and (ii) aminopyridine/pyrazine. There is no 3D structure for the enzyme in question and a limited number of structural diverse inhibitors. This project therefore entails the design of a homology model, MD optimization and VTS in a quest for novel inhibitors. At the moment, the project is at VTS stage. The results of this project will contribute greatly to the research community as Malaria remains without cure. The CHPC is being used for Molecular Dynamics and VTS screening of libraries. The project is being done at the University of KwaZulu Natal, Natural Products Research Lab.
Principal Investigator: Prof Pieter Rousseau
Institution Name: University of Cape Town
Active Member Count: 2
Allocation Start: 2021-10-05
Allocation End: 2022-04-03
Used Hours: 326894
Project Name: University of Cape Town Applied Thermal-Fluid Research Unit
Project Shortname: MECH1279
Discipline Name: Computational Mechanics
Rising energy demand and the imminent threat of climate change are critical issues in society today. Thermofluid systems provide the backbone of almost all energy conversion processes for renewable and conventional power generation, as well as heating and cooling systems such as heat pumps and refrigeration cycles. The Applied Thermofluid Process Modelling Research Unit (ATProM) at the University of Cape Town specialise in modelling these systems to evaluate novel technologies, improve the efficiency and control of processes, and detect anomalies for condition monitoring purposes.
Fundamental models are built using detail Computational Fluid Dynamics (CFD) and integrated one-dimensional thermofluid networks. These are often combined with advanced optimization techniques and data-driven models that are derived via artificial intelligence (AI) and machine learning (ML) techniques. This unique combination of fundamental thermofluid principles and AI techniques enable the development of versatile and accurate numerical tools to address industry needs. However, applying these techniques often involve resource intensive computational procedures. Given this, the services and facilities provided by CHPC are invaluable in generating industry-scale research outputs.
The current focus of research is on supercritical CO2 (sCO2) Concentrated Solar Power (CSP) plants, optimization of heat pump systems, flexibility of fossil fired power plants (coal, gas turbines and combined cycles), energy storage systems, biomass energy conversion and Physics Informed Neural Networks (PINN) applied to energy systems.
Principal Investigator: Prof Rotimi Sadiku
Institution Name: Tshwane University of Technology
Active Member Count: 10
Allocation Start: 2021-10-06
Allocation End: 2022-05-05
Used Hours: 206932
Project Name: NANOCOMPOSITE FOR ENERGY STORAGE
Project Shortname: MATS1376
Discipline Name: Material Science
Prof. Rotimi Sadiku, Prof. Suprakas S. Ray, Prof. Hamam, and Prof. Bonex oversee this research group of masters and doctorate students from several engineering specialties. The group's mission is to create high-performance energy storage systems. As a result, the research team will need to utilise High-Performance Computing resources to properly design and create such material. This is required in order to model and understand the transport characteristics of the metal ions in the anode, as well as how these ions organize inside the anode during charging and discharging. Experimental work will be prohibitively expensive without a thorough knowledge of the behavior of this material based on computer research. As a result, the atomistic approach will be used in this research.
Principal Investigator: Prof Chris Meyer
Institution Name: Stellenbosch University
Active Member Count: 5
Allocation Start: 2021-10-11
Allocation End: 2022-04-09
Used Hours: 78721
Project Name: Development of CFD models
Project Shortname: MECH1077
Discipline Name: Computational Mechanics
The research activity is led by Prof CJ Meyer from Stellenbosch University and is focused on developing computational models of industrial thermo- and fluid dynamic processes. There is a specific emphasis on the use of free software under the GNU General Public License. The industrial processes under consideration are those relevant to specific segments of industry in South Africa most notably power generation. Current projects include the simulation of large-scale air-cooled condensers as well as detailed simulations concerning the flow through axial flow fans used in the energy sector. The cohort of students are making excellent progress.
Principal Investigator: Prof Randhir Rawatlal
Institution Name: University of KwaZulu-Natal
Active Member Count: 1
Allocation Start: 2021-10-11
Allocation End: 2022-08-24
Used Hours: 22245
Project Name: CFD modelling of methane oxidation in monolith reactors
Project Shortname: MECH0894
Discipline Name: Chemical Engineering
The current project is conducted by the Systems Modelling and Data Analysis Research Group at UKZN and involves the development the "Nonlinear Response with Exogenous inputs (NRX)" data-driven modelling method for modelling of nonlinear dynamic systems from process data. The NRX method models the rate of change of an output variable with respect to its input variables and makes it possible for the optimal sampling of process data for modelling dynamic processes. The aim for developing this modelling method is achieve good predictive modelling accuracy for modelling of dynamic systems without needing a significant amount of data.
CHPC is currently being used to perform computer experiments that evaluate the predictive performance of the proposed method NRX in modelling a nonlinear continuous stirred tank reactor (CSTR) system and the Tennessee Eastman Process (TEP). Artificial Neural Networks are the model structure used to evaluate the performance of the NRX modelling method. The simulation experiments conducted for this work involve the following steps.
1. Simulating the CSTR and TEP Process to generate process data.
2. Optimal sampling of the data for these processes using the following methods:
a. Random Sampling
b. Latin Hypercube Sampling
c. Centroidal voronoi tessellation (CVT) sampling
3. Training an artificial neural network to model the CSTR and TEP Process using the various training datasets that have been sampled using the methods in step 2.
4. Evaluating the performance of the sampling methods and ANN models by predicting the CSRT and TEP process for an independent set of process inputs.
The CHPC has allowed us to scale the above simulation experiments. We are able to run by experiments for the multiple processes, multiple sampling methods and multiple dataset sizes simultaneously which has significantly improved the output and analysis of simulation results.
Principal Investigator: Prof Eno Ebenso
Institution Name: North-West University
Active Member Count: 6
Allocation Start: 2021-10-11
Allocation End: 2022-04-28
Used Hours: 139210
Project Name: Theoretical Investigations of Corrosion Inhibition Activities of Organic Compounds
Project Shortname: CHEM1176
Discipline Name: Chemistry
Our current research output on ANN modelling of organic corrosion inhibitors is geared towards revolutionizing the design and formulation of new corrosion inhibitors. The ANN modelling aids near-accurate prediction of inhibition efficiency of novel inhibitor material based on experimental and theoretical properties of previous related compounds. This model reduces experimental costs and speeds up the design of novel and efficient compounds that could serve as corrosion inhibitors. We have successfully developed ANN models for pyridazines (in an article published in Materials Today Communications) and quinoxalines (in an article published in Journal of Arabian Chemistry).
Principal Investigator: Prof Rosemary Dorrington
Institution Name: Rhodes University
Active Member Count: 8
Allocation Start: 2021-10-12
Allocation End: 2022-04-10
Used Hours: 10392
Project Name: Marine Natural Products Research
Project Shortname: CBBI0963
Discipline Name: Bioinformatics
Marine Natural Products Research Group, Department of Biochemistry and Microbiology, Rhodes University
The Marine Natural Products Research Platform covers the broad field of marine biodiscovery, focusing on the potential of bioactive secondary metabolites produced by marine organisms endemic to the Agulhas Bioregion as lead compounds for drug discovery.
Overview and aims
More than half of new pharmaceutical drugs on the market or in clinical trials are natural products and their derivatives and the majority of these are now coming from marine macrofauna and their associated microbiomes. The Maine Natural Produces research programme is active across the broad field of marine biodiscovery, focusing on exploring the potential of marine invertebrates and their associated microorganisms as sources of novel bioactive small molecules. Research projects include: (1) Biodiversity mapping of benthic habitats in the Agulhas bioregion, focusing on invertebrates, ascidians, soft corals and their associated microbiota; (2) Isolation and characterization of bioactive secondary metabolites of marine organisms; (3) screening marine natural product libraries for anti-viral, anti-cancer and antibacterial activity and (4) elucidating the metabolic pathways that result in the production of lead compounds to pave the way for engineering recombinant systems for their production for the pharmaceutical industry. Associated projects focus on the application of high throughput metagenomic and analytical chemistry technologies to map patterns and processes in marine ecosystems to provide tools for sustainable development of our ocean economy. All of these research activities require access to high performance computing facilities to curate, process and analyse large Next Generation Sequencing (NGS) datasets. Primarily, we have used the CHPC facilities to analyse NGS amplicon library data to characterise the microbial communities of both marine and terrestrial systems.
Principal Investigator: Prof Fourie Joubert
Institution Name: University of Pretoria
Active Member Count: 25
Allocation Start: 2021-10-12
Allocation End: 2022-04-28
Used Hours: 1620
Project Name: NGS Bioinformatics
Project Shortname: CBBI0905
Discipline Name: Bioinformatics
Tick infestations of cattle cause major health and economical problems. This project focuses on the discovery of molecules in ticks that may be modified to use as vaccines against ticks in cattle. The availability of the high performance computing infrastructure at the CHPC makes this research possible.
Principal Investigator: Dr Foster Mbaiwa
Institution Name: BIUST
Active Member Count: 2
Allocation Start: 2021-10-12
Allocation End: 2022-04-10
Used Hours: 4434
Project Name: Computational study of oxidative dehydrogenation of fatty acid methyl esters
Project Shortname: CHEM1461
Discipline Name: Chemistry
Professor Foster Mbaiwa's research group at the Botswana International University of Science and Technology is currently investigating how diesel produced from plant-based oils (like used sunflower oil), called fatty acid methyl esters, can be modified to resemble diesel from fossil fuels. This is important because diesel from plant oils, although it is environmentally friendlier is heavier than normal diesel and its continuous use on cars can damage the fuel pump. In our research we employ computational chemistry techniques to streamline the design of catalysts that can be used to convert the fatty acid methyl esters to fossil based diesel. Mainly we use the LAMMPS molecular dynamics code, thanks to the Center for High Performance Computing (CHPC), to simulate the reactions in the gas phase over a surface of a catalyst. These simulations are computationally demanding hence the need for high performance computers such as the one we use at the CHPC.
Principal Investigator: Dr Winfred Mulwa
Institution Name: Egerton University, Egerton, Kenya
Active Member Count: 10
Allocation Start: 2021-10-13
Allocation End: 2022-04-11
Used Hours: 140047
Project Name: Magnetic Refrigeration
Project Shortname: MATS1181
Discipline Name: Physics
Who: MATS1181 under PI Winfred Mulwa from Egerton University Kenya: Magnetic refrigeration. What: Magnetization and demagnetization of magnetic materials known as magnetic refrigeration that rely on magnetocaloric effect (MCE). In cooling technology, magnetic refrigeration which depends on magnetocaloric effect is commonly used in refrigerators to achieve exceptionally low temperatures. Why: . Magnetic refrigeration does not rely on the uses of harmful and ozone-depleting coolant gases. How: This work is done computationally. We use the Quantum Espresso code (Density Functional Theory). This purely depends on CHPC because all the calculations have to be done in CHPC. Thanks to CHPC. How is the project progressing: We have achieved all the objectives. The project is as planned. We are now working on what happens at the sink of the refrigerator (Thermoelectric properties)
Principal Investigator: Dr Jan Buys
Institution Name: University of Cape Town
Active Member Count: 9
Allocation Start: 2021-10-13
Allocation End: 2022-04-11
Used Hours: 603270
Project Name: Deep Learning for Low-resource Natural Language Processing
Project Shortname: CSCI1335
Discipline Name: Computer Science
The University of Cape Town Natural Language Processing group is a research group under the UCT Department of Computer Science that performs research on various topics related to the automated processing of language. The research group is led by Dr. Jan Buys and currently consists of 5 Masters students and 2 PhD students. Some undergraduate and honours students are also involved with the research.
The main focus is on developing methods for text generation in low-resource settings. Recent advances in Artificial Intelligence have been driven by the development of large neural networks that can "learn" how to perform various tasks by processing very large datasets - this is referred to as deep learning. For text generation, for example, deep learning systems have been developed that can automatically write realistic-looking news articles or other short pieces of text. However, deep learning techniques perform less well in settings where only small amounts of data are available. While text generation systems can perform well in English, similar systems for other South African languages (such as Nguni languages) are much less accurate or don't exist at all.
In this research project, we are performing foundational research to determine how we can adapt deep learning models to perform well in low-resource settings where only small amounts of data are available. Deep learning models are computationally intensive, even on small datasets, so they are trained on Graphical Processing Units to speed up the computations. We have developed language models, which are the models underlying deep learning text generation systems, for multiple South African languages. We are investigating multiple language model architectures, as well can multiple techniques for generating text using language models. Using the language models as basis we have developed novel techniques to automatically subdivide words into smaller units in order to enable modelling words in morphologically rich languages, which includes most South African languages. We have also developed systems to translate text from English into various South African languages, including both the Nguni and Sotho language groups, as well as systems to translate between different South African languages. We are investigating techniques to improve the performance given that only limited amounts of data (examples of translations) are available in these languages.
We have also developed language processing models for a number of other low-resource problems: Predicting clinical outcomes such as hospital readmission from doctor's notes; answering complex questions from documents; question answering in an interactive environment; and rewriting the style of text passages.
Principal Investigator: Mr Jean Pitot
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-10-14
Allocation End: 2022-05-19
Used Hours: 59315
Project Name: SAFFIRE Programme
Project Shortname: MECH1121
Discipline Name: Computational Mechanics
The University of KwaZulu-Natal's Aerospace Systems Research Group (ASReG) is at the forefront of rocket propulsion research in academia, both in South Africa and Africa as a whole. Located at the University's Discipline of Mechanical Engineering, ASReG primarily conducts research in the areas of liquid and hybrid rocket propulsion, as well sounding rocket development. The SAFFIRE Programme aims to develop South Africa's first fully-integrated liquid rocket engine, which is being designed for use in a small satellite launch vehicle. The long-term intention of the programme is for the engine to enable the indigenous development of a South African space launch capability. To meet its developmental goals, the programme requires the simulation of complex computational models to predict the fluid dynamic, thermochemical and structural performance of various engine components. These simulations are typically very computationally-expensive, and require extensive computational resources to run within a reasonable time frame. The CHPC offers the only realistic means of obtaining access to such resources. The programme is progressing well, and 2022 has seen the intake of numerous new postgraduate and undergraduate students. In total, 16 postgraduate students are currently involved in ASReG's cutting-edge projects.
Principal Investigator: Prof Stefan Ferreira
Institution Name: North-West University
Active Member Count: 5
Allocation Start: 2021-10-14
Allocation End: 2022-04-12
Used Hours: 75515
Project Name: Astrospheres
Project Shortname: ASTRO1277
Discipline Name: Astrophysics
In this project the astrospheric evolution of stellar wind cavities are simulated by executing numerical models on the CHPC architecture. Results from these are presented for different stars and also used as input into cosmic ray transport codes to simulate the background radiation in these cavities. These results are important in the search for habitable zones in other stellar systems.
Principal Investigator: Prof Edet Archibong
Institution Name: University of Namibia
Active Member Count: 7
Allocation Start: 2021-10-17
Allocation End: 2022-04-28
Used Hours: 152952
Project Name: (1) Computational Study of Semiconductor Clusters. (2) Computational Study of Bio-active Molecules Extracted from Plants
Project Shortname: CHEM0969
Discipline Name: Chemistry
Our group is the Computational Chemistry group at the University of Namibia, Windhoek, Namibia. We are using CHPC facilities, specifically the Lengau clusters to do our research. In one of our study, we are computing the properties of xerantholide, a compound with the potential of being used to treat gonorrhea. In another study, we are doping titanium oxide clusters with Cr, Mo, W in an effort to make the oxide more useful by absorbing in the visible region. The study also involves how the properties of the doped clusters vary with the length of the clusters.
Principal Investigator: Prof Cornie van Sittert
Institution Name: North-West University
Active Member Count: 20
Allocation Start: 2021-10-17
Allocation End: 2022-05-19
Used Hours: 5205300
Project Name: Computational Chemistry within LAMM at NWU (Potchefstroom)
Project Shortname: CHEM0778
Discipline Name: Chemistry
Since 1880 the main source of energy for South Africa was coal, and at present, coal provides 77% of South Africa's primary energy needs. However, electricity comes at a very high cost, namely air pollution and the influence of air pollution on human health. In 2004 the South African government reformed the legislation about air pollution. New alternative energy sources must be investigated to contribute to air quality management. These new alternative energy sources must be affordable, clean and renewable. Some of these alternative energy sources are fuel cells and batteries. Electricity from fuel cells and batteries is obtainable through electrochemistry, where free energy of spontaneous reactions is transformed into electricity. The hybrid sulphur (HyS) cycle is a potential replacement form of energy. The HyS cycle is a two-step water-splitting process that could be used to produce hydrogen. In the HyS cycle, the anode material used is platinum. However, because platinum is a rare and expensive metal, it is not an economically viable option for long-term and large-scale hydrogen production. Hence, various attempts to reduce or eliminate the platinum content while not compromising the process performance have been made. Investigations on a fundamental level are needed to understand the electrochemistry at the anode and make informed decisions on the type and amount of metal used in various platinum alloys. These types of investigations could be classified as computational chemistry. The resources needed for these investigations, namely various hardware and software, are vital. Although resources are available at NWU, it is not nearly enough to fully support the research covered in the Laboratory of Applied Molecular Modelling (LAMM). So if we, as researchers within the LAMM, did not have access to the CHPC resources, the progress of our research would be much slower.
Principal Investigator: Prof Marlo Moller
Institution Name: Stellenbosch University
Active Member Count: 9
Allocation Start: 2021-10-20
Allocation End: 2022-04-20
Used Hours: 4710
Project Name: Tuberculosis Host Genetics
Project Shortname: HEAL1360
Discipline Name: Health Sciences
Although approximately 25% of the world's population is infected with Mycobacterium tuberculosis, the vast majority will never develop any clinical tuberculosis (TB). Inter-individual variation in the immune response plays a major role in determining different clinical outcomes in infected persons. The Tuberculosis Host Genetics group based at Stellenbosch University aim to identify these unknown host genetic factors influencing immunity. Our investigations of the genetic contribution of the human host to individual and population susceptibility to tuberculosis span the full continuum of TB susceptibility, ranging from those extremely susceptible to disease to those seemingly resistant to infection. We have focused on TB genome-wide associations studies, host genes and strain interactions, the role of sex and ancestry in disease, TB "resisters", tuberculous meningitis and primary immunodeficiencies (PIDD) presenting with an increased TB susceptibility. Our experience in the genetic diagnosis of PIDD has allowed us to expand to the diagnosis of other rare genetic diseases by using the pipelines initially developed for PIDD. We also take the work to the cellular level by studying the function of the TB susceptibility genes discovered, with a specific interest in autophagy. Our insights into the latest genetic analyses available have allowed the work on genetic susceptibility to TB to benefit from population genetic analyses and next generation sequencing. Our work leverages the complex ancestry of South Africans, especially the unique genes from KhoeSan ancestry, to find novel genes and pathways involved in TB resistance or susceptibility, as shown by our admixture mapping, and pharmacogenetics. We supervised a ground-breaking project on the sequencing and de novo assembly of an admixed South African genome and establishing recombination maps to African groups. Our chief strength is on putting a uniquely African emphasis on the current highly computational field of genetics as it impacts on human disease, while exploring the downstream functional effects. The findings of our work could in future contribute to the development of host-directed therapies. Moreover, we are contributing to diagnosing rare diseases, including primary immunodeficiencies, in a developing country.
Principal Investigator: Prof Ken Craig
Institution Name: University of Pretoria
Active Member Count: 5
Allocation Start: 2021-10-21
Allocation End: 2022-04-21
Used Hours: 1097119
Project Name: CFD Modelling of Concentrated Solar Power Applications
Project Shortname: MECH0873
Discipline Name: Computational Mechanics
Central tower solar power plants are the most promising concentrated solar with research into new receivers and heat transfer fluids and collectors (heliostats) being ongoing world-wide. Researchers at the University of Pretoria's Clean Energy Research group under the leadership of Prof Ken Craig are using the resources at the CHPC to speed up calculations of both heliostat analysis (involving wind loads and fluid-structure interaction) as well as those required for a new point-focus central receiver design being developed. For heliostats, knowledge has been gained in predicting the peak loads that these reflectors will experience during all orientations and wind conditions. Ongoing work is focused on further reducing the cost of these Computational Fluid Dynamics (CFD) hybrid Large Eddy Simulation (LES):Reynolds-Averaged Navier-Stokes models to make them feasible on smaller-scale computers. This project has been ongoing since 2013 with the involvement of CHPC resources since 2016. The project has already delivered five masters students (Dawie Marais, Joshua Wolmarans, Derwalt Erasmus, Marcel Slootweg and Jesse Quick). PhD student Pierre Poulain is also nearing the end of this investigation into LES/RANS modelling of the atmospheric boundary layer and determining peak loads on solar collectors. For central receivers, work focused on the use of jet impingement, especially in the transient mode through either passive or active excitation. This phenomenon of heat transfer enhancement, although conceptually simple, it very hard to model accurately, and here the resources of the CHPC are invaluable, especially as LES is required to model the transient and 3D nature of swirling jet, the focus of the work of Jesse Quick. The jet impingement work is currently being taken further in another CPHC program for jet impingement boiling is used in electronics cooling.
Principal Investigator: Prof Emile Rugamika CHIMUSA
Institution Name: University of Kinsasha
Active Member Count: 18
Allocation Start: 2021-10-22
Allocation End: 2022-05-15
Used Hours: 318349
Project Name: Computational and statistical methodologies for human and environmental health prediction
Project Shortname: CBBI0818
Discipline Name: Bioinformatics
Computational and statistical methodologies for human and environmental health prediction research programme (CBBI0818) is running from the University of Cape Town, South Africa. This research programme is fundamentally aiming
1) to determinate the environment and genetics variation that cause human species to look different, having difference in allergy, drug responses and treatment.
2) to identify and quantify the pattern of inheritance of the observed trait, drug response and treatment variation among human.
These are addressed through the design of machine learnings and artificial intelligent methodologies and statistical approaches to analysis DNA data of thousands affected/unaffected subjects within a geographical region. In doing so, this programme will contribute to human health by increasing understanding of the genetic and environmental underpinnings of complex traits (or diseases), drug/treatment responses and drug/dosage responses and forecasting individual's health and traits such as the weather is forecasting today. The accumulation of experimental DNA data in Biology and new high throughput experiments are growing rapidly and give a huge number of data difficult to manage, to store and to analyse this existing and future dispersed information. Therefore, high-throughput technologies, such as next-generation sequencing, have turned molecular biology into a data-intensive discipline, requiring the field of bioinformatics to use high-performance computing resources and carry out data management and analysis tasks on large scale data. Today, the use of HPC has critically increased our researcher portfolio to meet the international standard with respect to large scale genomic era. This programme has already developed a number of genomic-based software tools that address African genetic variation challenges and provided advanced trainings around parallel programming with respect to large-scale genomic data to African postgraduate students/researchers
Principal Investigator: Prof Veikko Uahengo
Institution Name: University of Namibia
Active Member Count: 6
Allocation Start: 2021-10-22
Allocation End: 2022-06-15
Used Hours: 215977
Project Name: Solar Materials
Project Shortname: MATS1043
Discipline Name: Material Science
The University of Namibia-based research group is focusing on the development of Materials for energy. The group has been actively involved in this work for the past five years, having two PhD students on board. The research is centered around the pursuit for low-cost materials which can convert light photons into usable energy (in the form of electricity or other forms). The focus materials particularly in this project are on metal oxides (zinc oxide and copper oxide), minerals that are prevalent to mining in Namibia. Currently, these minerals are mostly exported in their raw forms, which has proved to have very little impact on economic returns. Thus, the main objectives of this research is focusing on value addition to local commodities such as minerals, 7in order to optimally increase the benefit of locals and common people. The objective is well in line with the SDGs of the United Nations towards Poverty Alleviation in communities. The second objective is on capacity building in the areas of materials chemistry, which is currently very low, to capacitate and bring awareness to the locals, as well as expanding the inclusive scope to reach out to women in this field (as the two PhD candidates in this project are women). In addition, given the soaring costs of chemicals and laboratory processes, computational methods have made life easier for researchers, when simulations can be performed first using computational methods, as a guiding platform to inform the experimental work. In the end it saves research groups from high costs of experimental trials and errors, thereby wasting costy chemicals and reagents. Thus, CHPC facility plays a very critical role here towards saving us in financial resources of buying unnecessary chemicals and reagents for experiments and also the precious time for Lab processes. The students in our groups have made some breakthrough in their research over the past years, which are now out published and accessible to the wider research communities. The other most important aspect of this research, was the scope of connectivity to different collaborators in the field of theoretical chemistry, as well as the introduction to CHPC some five years ago, the collaboration is so wide that now we can solve encountered problems with relative ease. The access of CHPC facility has greatly improved the efficiency and lives of our students and researchers in general.
Principal Investigator: Dr Eric Maluta
Institution Name: University of Venda
Active Member Count: 14
Allocation Start: 2021-10-22
Allocation End: 2022-05-12
Used Hours: 300061
Project Name: Energy Material Modeling in the Application of different Photovoltaic Technologies
Project Shortname: MATS0827
Discipline Name: Physics
The Department of Physics at the University of Venda is working on understanding the properties of different materials for application in renewable energy technologies and energy storage through the density functional theory simulations as performed under the CHPC cluster. The main aim is to reduce the cost of the material used in renewable energy technologies for the future use and mitigate climate change through energy supply. We are also building capacity of the rural students on the postgraduate studies and skill development. The CHPC is a gateway for the department of Physics of the University of Venda as an institution situated in the rural area to produce researchers and to publish research work. There is a growing number of students research work from honours to the PhD level. Several student are getting their masters and PhD degree through these initiatives.
Principal Investigator: Dr Lynndle Square
Institution Name: North-West University
Active Member Count: 5
Allocation Start: 2021-10-22
Allocation End: 2022-07-14
Used Hours: 148442
Project Name: Exploring poly(2,5)benzimidazole enhanced with carbon nanotubes for space applications
Project Shortname: MATS1088
Discipline Name: Physics
In this project, we are investigating a polymer nanocomposite for space applications. That has been enhanced with carbon nanotubes. The project itself has an experimental component and uses expertise across two institutions (NWU-UWC). We will explore ion bombardment, for which we will use a second national facility. The computational component for this project explores multiple scales of modelling, including fluid dynamics, Nano-scale modelling and molecular dynamics. The molecular dynamics will investigate the mechanical properties, which will be accompanied by experimental work and a second project will explore the surface ion interaction using LAMMPS/MedeA. This project relies on the Material Studio license provided by CHPC and the without the computing power it provides the project would have to reconsider its approach.
Principal Investigator: Dr Joseph Mutemi
Institution Name: University of Nairobi
Active Member Count: 5
Allocation Start: 2021-10-27
Allocation End: 2022-05-12
Used Hours: 5346
Project Name: Numerical weather and climate modeling, prediction, forecasting and change projections for Africa and sub regions
Project Shortname: ERTH1131
Discipline Name: Earth Sciences
The ERTH1131 Research Group focuses on research themes of interest over East Africa in terms of model evaluation and efforts to improve operation seasonal to sub-seasonal forecasting. During the reporting period, the researchers have utilized the state-of-the-art climate models to evaluate different climate processes in East Africa. This includes characterization of thermodynamic processes related with the Turkana Jet, a phenomenon that is responsible for the wind power potential over the northern Kenya. In addition to be a resource, the Turkana Jet is also responsible for the aridity in the region. Proper representation of the system in regional and global models is therefore useful for both resource and hazard monitoring in Kenya and East Africa. Our results indicate that the poor representation of the intensity of the Turkana jet is responsible for the rainfall biases in climate models over East Africa. Other results on land-atmosphere interactions indicated that global models over-estimate the strength of the land-atmosphere coupling in parts of East Africa and the Sahel. In this regard, with drying trends of soil moisture observed in global scales, this is regions of strong coupling are thereby more vulnerable to increase in aridity. Proper intervention measures, especially related with water conservation, and activities that replenish soil moisture, are key adaptation measures that may slow down aridity in East Africa
Principal Investigator: Prof Ponnadurai Ramasami
Institution Name: University of Mauritius
Active Member Count: 5
Allocation Start: 2021-10-27
Allocation End: 2022-05-05
Used Hours: 24153
Project Name: Computational Chemistry Methods to Study Structural and Spectroscopic Parameters
Project Shortname: CHEM1290
Discipline Name: Chemistry
I am the group leader of the Computational Chemistry Group in the Department of Chemistry of the University of Mauritius. The research interest of the group ranges from fundamental to applied chemistry. We are also collaborating with experimentalists, mainly organic and inorganic chemists from African countries. For the time being we are using Gaussian and ORCA. We have two publications for the period Oct 2021-Apr 2022 and one of them has been selected as 2022 RSC Advances hot articles. Without the use of the CHPC facility, it would not have been possible to carry out high level computations. This facility is allowing us to aim at carrying out high level research and to eventually have good publications in high impact factor journals.
Principal Investigator: Dr Pedro Monteiro
Institution Name: Council for Scientific and Industrial Research
Active Member Count: 17
Allocation Start: 2021-10-29
Allocation End: 2022-04-29
Used Hours: 516848
Project Name: Southern Ocean Carbon - Climate Observatory SOCCO
Project Shortname: ERTH0834
Discipline Name: Earth Sciences
SOCCO is a CSIR-led and DSI co-funded programme that uses the comparative geographical advantage that South Africa has in the Southern Hemisphere to contribute to the global research challenge of understanding the role of the Southern Ocean in global and regional climate. The SOCCO research niche is to explore the role played by fine scale ocean dynamics in improving the climate sensitivity of ocean and coupled numerical models with a special focus in understanding the Southern Ocean's role in the CarbonClimate links. To address the scales of dynamics in the ocean, we primarily use the coupled NEMO-PISCES model in a hierarchy of resolutions from 200 km to 2 km - using the models as experimental platforms. The finer the resolution, the greater the number processes that contribute to ocean-atmosphere carbon exchange. This understanding contributes to reducing model biases in process models and climate projections as well as the development of a South African Earth Systems Model (ESM). This work advances research in global climate change based on the importance of feedbacks of the Southern Ocean dynamics on the carbon-climate system. Additionally, the ESM will be used to explore how the climate sensitivities of the Southern Ocean evolve over the 21st century as well as being South Africa's first submission to the 6th Assessment Report of the IPCC. ESM will be the basis for the medium to long term climate projections in Africa which will then support societal needs in health, water resource planning, agriculture, biodiversity management and more broadly in economic development in the region. From regional, forced ocean models to fully-coupled earth system models comprising of interacting ocean, ice, atmosphere, land, ocean biogeochemical models, numerical modelling of earth system components require massively parallel computing resources. The size and complexity of these models are such, that for SOCCO's work, the CHPC is indispensable for the resources and service it provides.
Principal Investigator: Prof Malik Maaza
Institution Name: University of South Africa
Active Member Count: 4
Allocation Start: 2021-11-03
Allocation End: 2022-06-08
Used Hours: 238603
Project Name: Modelling of Functional Materials at the Nanoscale
Project Shortname: MATS1306
Discipline Name: Physics
The UNESCO UNISA ITL-NRF Africa Chair in Nanosciences & Nanotechnology (U2ACN2), a trilateral partnership between the UNESCO UNISA & iThemba LABS-National Research Foundation of South Africa, has established a multidisciplinary research program in materials at the nanoscale. The multidisciplinary approach of the U2ACN2 chair cements the several fragmented Africa national efforts in nanosciences and nanotechnology and addresses urgent societal needs in the water, energy, and health sectors in Africa.
Computational research uses complex models in various ways, all of which advance materials science and engineering. These computational models can help researchers understand the outcome of an experiment, identify the most promising avenues for future experiments, and give us insight into processes that cannot be easily explored in the lab. The U2ACN2 center thanks CHPC as an outstanding High-Performance Computing center, which assists our researchers by providing access to computational resources necessary to construct, analyze, and interpret their complex data in the field of nano and materials science.
Our recent projects have been focused on finding an approach via simulation methods including DFT & MD to use the coated metal nanoparticles as deliverer with proposed drugs to treat diseases caused by the coronavirus and Malaria insect.
The Modelling computational publication on SARS-COV2 in Nature Scientific Report 2021 has been loaded & publicized on various institutional websites, UNISA, iThemba LABS & NRF. Also, the same publications are now on the open-access platform of Springer & Nature Publishing houses. Likewise, it is intended to use the same approach to investigate if ivermectin & Artemisinin phytoactive compounds can bioconjugate with nanoscaled metallic nanoparticles. The target is to study the efficacy of such bioconjugated nanosystems against Malaria.
Principal Investigator: Prof Ignacy Cukrowski
Institution Name: University of Pretoria
Active Member Count: 4
Allocation Start: 2021-11-03
Allocation End: 2022-05-12
Used Hours: 35748
Project Name: Understanding chemistry from QM study of molecular systems
Project Shortname: CHEM0897
Discipline Name: Chemistry
'Modern' chemistry is a very well established science for at least 150 years. Imagine life without CHEMISTRY – simply impossible. Most items we touch are synthetic or treated by synthetic materials (paints, varnish, waxes, etc.), technologies strive on chemical principles and developments, and energy related projects and resources (solar and fuel batteries, nanomaterials) are all chemistry related, to mention just a few examples. Availability of powerful computers and PCs changed everything and particularly Science as it provides a platform to explore and challenge our orthodox concepts on how the universe of chemistry works. A chemical bond is the central concept in chemistry. In the orthodox chemistry, a compound is a set of atoms that, according to the Lewis over 100 year's old dogma, are covalently bonded. All academic textbooks as well as many software applications have incorporated that dogma that often fails to explain formation and properties of compounds.
Prof Cukrowski research group in the Department of Chemistry, University of Pretoria is involved in fundamental studies focused on chemical bonding for several years now. We interpret classical chemical bonding, regardless of its kind, as an interaction governed by universal laws of physics. We expended the concept of a classical 2-atom chemical bond to polyatomic interactions or even all atoms of a molecular system. To this effect, a unified molecular-wide and electron density based concept of chemical bonding was proposed recently where, in principle, are no classical chemical bonds, as we know them. This concept is being extensively tested on numerous molecular systems investigated by postgraduate students in modelling e.g., reaction mechanisms to understand how and why new compounds are or are not formed as planned. We are using CHPC facilities and one must give lots of credit to the dedicated staff who strives to maintain this national facility.
Principal Investigator: Other Christo Venter
Institution Name: North-West University
Active Member Count: 6
Allocation Start: 2021-11-04
Allocation End: 2022-05-05
Used Hours: 38507
Project Name: Astrophysical Calculations and Data Analysis
Project Shortname: ASTR1245
Discipline Name: Astrophysics
Who? Several astrophysicists from the North-West University, along with international collaborators.
What? We are studying pulsars via complex numerical models and multi-wavelength data fitting.
Why? Pulsars have been an enigma since their discovery in 1967. Much progress has been made over the past decades, with the latest breakthroughs coming from the Fermi Large Area Telescope and ground-based Cherenkov telescopes that detected pulsars in the GeV to TeV energy range. We will continue to use the latest high-energy data to constrain our cutting-edge models and thus improve our understanding of pulsars. Broadly speaking, this contributes to highly skilled graduates who can make an impact in industry as well as communicate basic science to the public.
How? We are running large parallel codes on the CHPC in order to compute and fit models to data.
How is the project progressing? The progress has been steady and several publications are in view.
Principal Investigator: Dr Ndumiso Mhlongo
Institution Name: University of KwaZulu-Natal
Active Member Count: 10
Allocation Start: 2021-11-08
Allocation End: 2022-05-09
Used Hours: 87435
Project Name: Biomolecular Modelling Research Unit
Project Shortname: HEAL1002
Discipline Name: Health Sciences
Biomolecular Modelling Research Unit is located at the University of KwaZulu-Natal, Howard College Campus and headed by Dr Ndumiso N. Mhlongo. This research group employs CHPC resources to conduct molecular dynamics simulations to analyse the dynamical properties of macro-molecules such as folding and allosteric regulations; ligand-receptor interactions; atom-to-atom interactions; protein-protein interaction;virtual screening and molecular docking of active molecules; QSAR etc, for the design and development of inhibitors against diseases and disorders. Through the use of CHPC resources, we recently identified accolate, sorafenib, mefloquine, and loperamide as potential inhibitors of M. tuberculosis by through a drug repurposing principle. These FDA-approved drugs are currently used to treat malaria, liver & kidney cancer, asthma and irritable bowel symptoms. These drugs will now be further evaluated to validate their therapeutic potential against TB through in vitro/in vivo experimental methods.
Principal Investigator: Dr Nikiwe Mhlanga
Institution Name: Mintek
Active Member Count: 7
Allocation Start: 2021-11-08
Allocation End: 2022-05-09
Used Hours: 494143
Project Name: Hydrogen Storage Materials and Catalysis
Project Shortname: MATS0799
Discipline Name: Material Science
Advanced Materials Division (AMD) houses several research and development groups; catalysis, nanotechnology (health, water purification, and sensors), and physical metallurgy. The nanotechnology platform develops diagnostic kits for diseases such as malaria, tuberculosis. The kits are in the form of field-based quantitative lateral flows and quantitative lab-based surface-enhanced Raman spectroscopy (SERS) biosensors. Understanding the precursor materials and their interaction with the analyte is of paramount importance. The HPC platform enables simulations/calculations of these systems (large systems) to gain insight into their chemistry and biology for the fabrication of better kits. Also, HPC enables the study and establishment of new memory-shaped alloys by the physical metallurgy group. Ultimately the properties inform the experimental development of the alloys envisioned for aerospace engines. The development of diagnostic kits is one of the key factors in the eradication of illnesses such as TB and malaria. And the end product is envisioned to benefit the clinical sector. The structures normally simulated for these projects are massive and require a detailed tighter computational setting to yield informative outputs. We do not have the capacity to run such calculations on our local computer hence the need for the HPC platform which we continue to benefit from.
Principal Investigator: Prof Jeanet Conradie
Institution Name: University of the Free State
Active Member Count: 8
Allocation Start: 2021-11-08
Allocation End: 2022-05-09
Used Hours: 2449311
Project Name: Computational chemistry of transition metal complexes
Project Shortname: CHEM0947
Discipline Name: Chemistry
The CHPC resources made it possible for my research group to understand and shed light on experimental observation. The theoretical calculated results also make it possible to predict experimental behaviour. For example, it was possible to get the correct DFT approach to support the experimentally observed oxidative addition reaction of a model rhodium catalyst, like the rhodium catalysts being used in the Monsanto process. The Monsanto process is among the most popular examples of an industrial process that is catalysed by metal complex solution, by converting methanol into acetic acid (Polyhedron 199 (2021) 115095; Data in Brief 35 (2021) 106929; Molecules 2022 27 1931). It was also possible to show that theoretical DFT techniques can be used to accurately simulate complex properties of a series of metal-to-ligand charge transfer complexes of row-4 (Mn, Fe & Co) and column-8 (Fe, Ru & Os), with application in electro- and photocatalytic applications involving reduction of CO2 and H2O, and dye-sensitized solar cells (Electrochemistry Communications 136 (2022) 107225).
Principal Investigator: Dr Caroline Kwawu
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 7
Allocation Start: 2021-11-08
Allocation End: 2022-05-09
Used Hours: 249845
Project Name: Renewable Energy Materials
Project Shortname: MATS1146
Discipline Name: Material Science
The Kwawu's materials catalysis group carries out research in the area of renewable energy materials. We are researchers in the Chemistry Department of the Kwawe Nkrumah University of Science and Technology, with focus on computational materials chemistry. We study the electronic, structural and electronic properties of materials and predict improved materials for applications in supercapacitors, batteries, electrolyzers, gasifiers, photoelectronic and solar cells. As populations grow the need for energy is expanding requiring expansion of our energy sources.
Our research is critical to assist develop cheaper and efficient materials for renewable energy generation, and to make energy affordable and accessible to all.
We employ computational chemistry approach to finding energy solutions, whereby we employ mathematical theories implemented in computer algorithms to understand chemical systems. The CHPC provides computer hardware of high storage and memory space to solve complicated equations.
So far, we have been able to predict more efficient Fe and Cu based materials for CO2 conversion and also contributed to understand plastic waste and agricultural waste gasification into electricity. We are still studying how we can improve metal organic frameworks and perovskites as organic and inorganic materials for solar energy conversion into electricity. We are also looking at porous carbon materials as electrodes in supercapacitors and batteries to ensure electricity storage for future use.
Principal Investigator: Prof Giuseppe Pellicane
Institution Name: University of KwaZulu-Natal
Active Member Count: 7
Allocation Start: 2021-11-08
Allocation End: 2022-05-09
Used Hours: 577216
Project Name: Computational Study of Structure-Property Relationships in polymer blends relevant to OPV devices
Project Shortname: MATS0887
Discipline Name: Physics
Polymers relevant to organic photovoltaics and surface engineering are our focus in this research project. The group leader is Dr G. Pellicane, who is a (honorary) associate professor at UKZN and NRF C1 rated scientist. Members include Mr Emmanuel Ayo-Ojo and Mr. S. Mamba (PhD students), Prof. Tsige (Full Professor of Polymer Science at the University of Akron, US) and Prof. Workineh (Associate Professor). We work in the fields of theory and computer simulation to study the microscopic behaviour of complex fluids, with density functional theory and molecular dynamics simulations. We published several articles in scientific journals within the framework of this project, and presented our results in international conferences/workshops, that will allow for sharing and diffusion of vital knowledge with the international scientific community. We value much the skillful and resourceful staff members at CHPC (Dr Anton Lopis), and are grateful to CHPC for the generous allocation of computational resources granted to us.
Principal Investigator: Prof Johan du Preez
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-11-08
Allocation End: 2022-05-09
Used Hours: 24398
Project Name: Advanced inference and learning in structured probabilistic machine learning systems
Project Shortname: CSCI1419
Discipline Name: Electrical Engineering
Prof Johan du Preez from Stellenbosch University leads a research group with a focus on advanced inference and learning in structured probabilistic machine learning systems. We are interested in large probabilistic machine learning systems expressed as a blending of graphical and/or neural models. Inference and learning lean towards Bayesian systems incorporating variational approaches in an open-set semi-supervised context. The group uses an in-house probabilistic graphical model toolkit named EMDW. EMDW is used to construct probabilistic machine learning models that are highly modular and fast. The models have found application in speech recognition, image recognition, logic puzzles and many other areas. The use of probabilistic graphical models for larger problems often requires a lot of resources. The CHPC resources have allowed us to successfully scale our models to handle larger problems by using multiple cores and large amounts of memory. A new model was introduced and tested on the CHPC, Markov Chain Latent Dirichlet Allocation, and the results thereof was submitted to Interspeech 2022.
Principal Investigator: Dr Victoria Nembaware
Institution Name: University of Cape Town
Active Member Count: 15
Allocation Start: 2021-11-13
Allocation End: 2022-06-09
Used Hours: 71330
Project Name: Sickle Africa Data Coordinating Center (SADaCC)
Project Shortname: CBBI1243
Discipline Name: Health Sciences
The GeneMap group has made several discoveries in the last couple months which would not have been possible without using the CHPC services and resources. These discoveries include the identification of novel genes for Hearing Loss from African genomic data; understanding the evolution and origins of mutations associated with Hearing Loss. In addition, the group has developed a multi-national Sickle Cell Disease registry and research towards the understanding of high blood pressure in Sickle Cell Disease patients was also conducted.
Principal Investigator: Prof Hadley Clayton
Institution Name: University of South Africa
Active Member Count: 4
Allocation Start: 2021-11-14
Allocation End: 2022-05-15
Used Hours: 100246
Project Name: Bioorganometallic Chemistry of Transition Metals
Project Shortname: CHEM1288
Discipline Name: Chemistry
Clayton Organometallic Research Group at the University of South Africa. Computer applications will be used to investigate he bioorganometallic chemistry of platinum group metals (PGMs) and their derivatives as potential metal-based anticancer drugs. Density functional methods will be applied to investigate chemical and physical properties of new transition metal complexes synthesized. Molecular docking studies are applied to study metal complex-protein interactions.
Principal Investigator: Dr Nangamso Nyangiwe
Institution Name: Tshwane University of Technology
Active Member Count: 2
Allocation Start: 2021-11-15
Allocation End: 2022-07-28
Used Hours: 63442
Project Name: Application of density functional theory in engineered nanoparticles
Project Shortname: MATS1427
Discipline Name: Material Science
I recently started a research group at TUT Physics department. My research field is on computational materials modelling, I mostly use density functional theory (DFT) and molecular dynamic (MD) to model and predict different properties of materials. Since its arrival, the density functional theory (DFT) has experienced a substantial development, establishing nowadays as the most employed tool to tackle many different facets of the electronic structure of molecular systems. The density functional theory is, in principle, an exact method. It works with functionals, which give the energy or other properties in terms of the electron density n(r). Only a small but essential contribution to the energy, the exchange-correlation (XC) energy, thus far remains unknown as a functional of the density and has to be approximated. The overall objective of our projects is to model and predict different properties of materials so that we can assist experimentalists when they are designing their experiments.
Principal Investigator: Dr Joseph Simfukwe
Institution Name: Copperbelt University
Active Member Count: 1
Allocation Start: 2021-11-15
Allocation End: 2022-05-19
Used Hours: 4735
Project Name: Computational Materials Science Research Group CBU
Project Shortname: MATS1469
Discipline Name: Physics
Computational Materials Sciences Research Group CBU is based at the Copperbelt University (CBU) in Zambia headed by Dr. Joseph Simfukwe a PhD graduate from University of Pretoria (UP). It was established soon after Dr. Simfukwe completed his studies at UP and returned to CBU. The group currently works in collaborations with Dr. Mapasha and Prof. Diale of UP and other partners within Africa and abroad. There is an urgent call world over to reduce the continuous emission of greenhouse gasses such as carbon dioxide (CO2) from burning fossil fuels (FFs). The reliance on fossil fuels to supply energy for human use and industrialization has led to the increase in the earth's surface temperature and negatively affecting human health and sustainability. Therefore, it is apparent, that clean, renewable, cheap and sustainable sources of energy must be found to replace the FFs while meeting the increasing energy demand. Renewable energy sources such as solar have the potential to meet the global energy demand sustainably. To harness energy from the Sun, we need an efficient and inexpensive route to convert solar energy into chemical energy such as hydrogen (H2) fuel. H2 is envisioned a primary media for the storage and distribution of the harvested energy from the Sun. It is also an environmentally friendly fuel because only water is produced during combustion. One promising route for harvesting H2 is photo-electrochemical (PEC) water splitting. It involves the use of suitable semiconductors to decompose water molecules into constituent H2 and O2 upon photon absorption.
In this project, our focus is to study different properties of materials that can be used in PEC water splitting and their applications in addressing various societal needs such as clean energy and its storage. The method of study involves the application of first-principles simulations based on state-of-the-art density functional theory (DFT). This is a computational study which heavily depends on computational resources from the CHPC clusters. Computer simulations based on quantum mechanical methods have become important in understanding various properties of matter at both atomic and molecular levels. Electronic structure calculations methods can accurately predict various physical properties of materials and also provide a detailed understanding of experimental results. We greatly appreciate will the support we receive from CHPC for cluster resources.
Principal Investigator: Dr Evans Benecha
Institution Name: University of South Africa
Active Member Count: 11
Allocation Start: 2021-11-16
Allocation End: 2022-05-17
Used Hours: 288628
Project Name: Defect engineering in novel 2-D materials
Project Shortname: MATS1025
Discipline Name: Physics
The research programme titled "Defect engineering in novel 2-D materials" focuses on investigation of the properties of defects in single layered materials - usually one atom thick, using Density Functional Theory (DFT) calculations. The group spearheading this research is headed Dr Evans Benecha from the University of South Africa, and it consists of senior researchers, postdocs and graduate students. Single atom layered materials possess interesting properties as opposed to many atom thick materials. For example, Graphene – which a single layer of carbon atoms –enables electrons to flow much faster than silicon, is the world's strongest material, harder than diamond, yet lightweight and flexible. This unique combination of superior properties makes it a credible material for new technologies in a wide range of fields. Some of the envisaged applications of Graphene as well as other single atom layered materials include making of medical implants, solar cells, stronger sports equipment, and batteries; just to name a few. However, the properties of these materials are not yet well understood for these applications to be realized. Dr Benecha and his group's approach towards understanding these materials involves use of large computational resources, and access to the CHPC computing facility has significantly reduced the time needed to generate and analyze data. The results from this project will not only assist in the current understanding of the role of defects in these materials, but will also help to reveal potential new functionalities.
Principal Investigator: Dr Francis Opoku
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 15
Allocation Start: 2021-11-16
Allocation End: 2022-05-17
Used Hours: 775719
Project Name: High-Throughput theoretical design of two-dimensional materials as an ultrasensitive toxic gas sensor
Project Shortname: MATS1423
Discipline Name: Chemistry
The release of anthropogenic pollutants toxic gases into the atmosphere and environment is a global concern. The use of sensor technologies raises significant legal concerns with privacy advocates and the industrial world are not immune to these changes. Currently, the world is in the very early stages of the Fourth Industrial Revolution (4IR), where new technologies are transforming manufacturing and making factories 'smart'. The underlying features of the reaction processes, such as the structural parameters and adsorption energy of the adsorbed volatile organic compounds, toxic gas and drug residues in aqueous solution, remain uncertain and such information cannot be directly achieved from experimental work. Thus, the fundamental physical driving forces, which control the reactivity of volatile organic compounds, toxic gas species and drug residues with two-dimensional-based van der Waals heterostructure surface are still poorly understood. To address these issues, computational simulations are indispensable to offer fundamental insights to further advance the current state of knowledge. The computational laboratories of the Department of Chemistry Department and the Centre for High Performance Computing, Cape Town are equipped with most of the advanced computational software and hardware required for this study. The accumulated knowledge will be shared with the scientific community through meetings with collaborators, publications in international peer-reviewed journals and presentations at national and international conferences and workshops. It will also offer an opportunity for training and innovation as more engineering and science graduates could be involved in the production of these sensors, while those with a business background will be engaged in the distribution and marketing.
Principal Investigator: Dr Parvesh Singh
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2021-11-16
Allocation End: 2022-05-17
Used Hours: 22821
Project Name: Molecular dynamics and docking studies of organic compounds
Project Shortname: CHEM0795
Discipline Name: Chemistry
I am working as an Associate Professor in the School of Chemistry at UKZN. My research focuses on designing and synthesis of heterocyclic compounds of medicinal relevance. We are using different computer-based technologies such as molecular docking, QSAR, QSPR, and pharmacophore modeling to predict novel chemical assemblies from the online databases. The identified lead molecules are synthesized in our organic laboratory and tested in vitro. All the computer simulations are performed on the CHPC cluster. The CHPC facility is an excellent platform that helps us as researchers in SA to run computer simulations and use the compiled data to substantiate our experimental results, which in turn enable us to publish our research in high-impact journals. Running these jobs on normal computers would have taken months to finish or would die in middle. With this supercomputing facility in our hand, we not only managed to run complex calculations but obtained constructive scientific explanations for our experimental results.
Principal Investigator: Prof Rudolf Laubscher
Institution Name: University of Johannesburg
Active Member Count: 3
Allocation Start: 2021-11-17
Allocation End: 2022-05-18
Used Hours: 3099
Project Name: Numerical modelling of Machining
Project Shortname: MECH1249
Discipline Name: Computational Mechanics
This work is run, under the titanium machining research group of Professor RF Laubscher from the Mechanical Engineering Science department of the University of Johannesburg.
The hole-drilling technique for measuring near-surface residual stresses requires the use of a numerical method, such as the finite element method to determine the surface strain-to-stress calibration coefficient for each drilled hole-depth increment.
Due to the large plastic deformation and temperature gradient generated from the machining of a part, it is critical to measure the near-surface residual stresses. These residual stresses can affect the fatigue life of the component, depending on the direction and magnitude thereof.
The incremental hole-drilling technique for the measurement of residual stresses is suited to this application if applied with stringent practices.
The measurement of the residual stresses in machined titanium alloys then assists in the development of a numerical model to "predict" the near-surface residual stresses, which are less expensive and time-consuming than measurement work.
The project requires the use of the CSIR CHPC for the large number of smaller simulations needed for the "a" and "b" calibration coefficients.
The CSIR CHPC resources are required for the large "machining simulations". These simulations are taking approximately 72 hrs on a single 6-core 32GB RAM Linux desktop PC. There are 64 of these runs to be conducted. We are hoping to install and use MSC Marc with LUNAR for the design of experiments work.
Additional computation resources are essential in this regard.
Principal Investigator: Dr Francois van Heerden
Institution Name: Nuclear Energy Cooperation of SA
Active Member Count: 8
Allocation Start: 2021-11-17
Allocation End: 2022-06-30
Used Hours: 22901
Project Name: Research Reactor Fuel Burnup and Material Activation
Project Shortname: INDY0788
Discipline Name: Physics
The Radiation and Reactor Theory (RRT) group at the South African Nuclear Energy Corporation (Necsa) provides calculational support for the SAFARI-1 research reactor, and perform research in the general area of nuclear analysis. We also develop and maintain our own suite of reactor simulation software, called OSCAR. This suite is the primary operational support tool for the SAFARI-1 reactor, and is used at a number of international facilities, which range from small university based research reactors, to larger national facilities. We are also involved in a number of calculational benchmarking exercises, facilitated by the International Atomic Energy Agency (IAEA). High Performance Computing plays an important role in the development of multi-physics high fidelity modelling of nuclear reactors, as conducted in the OSCAR-5 code suite. During the past year the work performed in conjunction with the CHPC has notably contributed the better methods and improved maturity of the system, as well as to journal and international peer-reviewed conference publications related to the OSCAR system. Necsa would like to thank the CHPC for the excellent service during the year, and ongoing support in furthering the development of high-level local IP.
Principal Investigator: Dr Rendani Mbuvha
Institution Name: University of the Witwatersrand
Active Member Count: 2
Allocation Start: 2021-11-18
Allocation End: 2022-05-26
Used Hours: 38075
Project Name: Bayesian Methods for Neural Networks
Project Shortname: CSCI1332
Discipline Name: Data Science
We are part of a research group that is led by Prof Tshilidzi Marwala at the Universities of Johannesburg and Witwatersrand. The group focuses on Bayesian methods in machine learning. Bayesian methods allow for a principled understanding of the uncertainty around the parameters of complex methods like neural networks. Bayesian inference frequently requires the use of computationally intensive methods such as Markov chain Monte Carlo - which require CHPC resources to accelerate experimental time through extensive parallelization. The work has resulted in two publications as well as numerous works in progress including an accepted book.
Principal Investigator: Prof Tobi Louw
Institution Name: Stellenbosch University
Active Member Count: 4
Allocation Start: 2021-11-18
Allocation End: 2022-05-19
Used Hours: 64484
Project Name: SU Process monitoring, modelling and control
Project Shortname: CSCI1370
Discipline Name: Applied and Computational Mathematics
The Stellenbosch University Process Monitoring, Modelling and Control group aims to use recent advances in computational power and "big data" to improve the way in which industrial chemical- and mineral processing plants run. We believe that the wealth of data available from industrial plants can be used to significantly improve their profitability. An exciting new development that we are looking forward to explore further is the use of neural networks for process control, specifically within the reinforcement learning framework. The CHPC is essential to exploring this avenue further, given the computational costs of running multiple simulations to train neural networks.
Principal Investigator: Dr Kiprono Kiptiemoi Korir
Institution Name: Moi University, Eldoret, Kenya
Active Member Count: 9
Allocation Start: 2021-11-18
Allocation End: 2022-05-26
Used Hours: 578106
Project Name: Thermal characterization of MoS2 nanostructures: an ab initio study
Project Shortname: MATS0868
Discipline Name: Material Science
This research group draws its members from Moi University, Computational Material Science Group (CMSG), in addition, we have collaborators: Dr. Haffad of University of Beijaia- Algeria, and Dr. Re Fiorentin of Istituto Italiano di Tecnologia (IIT) Italy. Our research activities focus mainly on material discovery and optimization for energy, optoelectronics, and ultra-hard industry. The predictive techniques employed in these studies can hasten the development of robust and cost effective materials for energy such as electrodes for hydrogen fuel production via water splitting, and materials for ultra-hardness applications, which is anticipated to have a positive impact on the society by promoting environmental conservation, thus merit the utilization of public resources.
These predictive approaches requires huge computational effort as it entails solving mathematical formulas, such as Schrödinger and Newtonian equations. For this reason, availability of the state of the art High Performance Computing facility, such as CHPC is a critical component for implementation of this work. Solving Schrödinger and Newtonian equations can yield various properties of interest such as structural, mechanical, electronic and optical properties that are essential for comprehensive characterization of these systems.
For example, in our recent work, using ab initio Density Functional Theory calculations, we explored the impact of anionic vacancies on the mechanical properties of NbC and NbN. Our results showed that the presence of anionic vacancy defects in NbC and NbN tend to deteriorate the mechanical properties and ultimately the mechanical hardness due to vacancy softening, which can be attributed to defect induced covalent to metallic bond transition.
Further, it was observed that anionic vacancy defects in NbC tend to modify the mechanical toughness; in particular, NbC becomes ductile in the presences of carbon vacancy defects dosage of up to 6% while the toughness NbN was found to insensitive to nitrogen defects concentration of upto 8%. Consequently, stringent control of anionic defects during synthesis of NbC and NbN is critical for realization of the desired mechanical response that can make these materials ideal for super-hard and related applications.
Principal Investigator: Dr Hezekiel Kumalo
Institution Name: University of KwaZulu-Natal
Active Member Count: 11
Allocation Start: 2021-11-22
Allocation End: 2022-05-23
Used Hours: 254148
Project Name: Molecular modeling and computer aided drug design
Project Shortname: HEAL1009
Discipline Name: Health Sciences
The group covers a wide range of research areas in computational and molecular modelling, with a focus on biological systems and approaches to drug design.
The main interest is in the design and investigation of biologically and therapeutically oriented targets. This is achieved by applying computational methods to the study of chemical and biochemical reactivity problems, with particular emphasis on the transition state, environmental influences on mechanisms, the origins of catalysis and the interpretation of kinetic isotope effects. This includes mechanistic pathways and transition states for reactions in enzymes and solutions, the development of enzyme inhibitors and the exploration of the binding and catalytic theme of evolved targets, and the application of advanced computational approaches to understanding protein structure and function. We are involved in projects such as understanding drug resistance mechanisms using various computational tools QM / MM MD simulations Quantitative Structure Activity Relationship (QSAR) conformational analysis of biomolecules Bioinformatics tools Applications Development projects Approaches to improve the results of binding free energy calculations Development of biomolecule parameters: ongoing projects Software implementations: ongoing projects. All the projects we undertake in the group contribute to the training of pharmaceutical scientists with specific skills for both the pharmaceutical industry and academia. This has the potential to stimulate the local pharmaceutical industry in South Africa to produce high quality and affordable medicines for optimal and cost-effective patient care, rather than depending on multinational pharmaceutical companies. We are still a young group, so funding is an issue. The resources of the CHPC allow us to develop hypotheses that we can then test experimentally much faster and with higher throughput than if we were to develop the hypotheses we want to test through experiments alone. It has given us the opportunity to explain experimental data that we could not easily explain, such as the binding landscapes of different enzymes and the mechanisms of action of different inhibitors.
Principal Investigator: Dr George Manyali
Institution Name: Masinde Muliro University of Science and Technology, Kakamega, Kenya
Active Member Count: 13
Allocation Start: 2021-11-22
Allocation End: 2022-06-02
Used Hours: 310745
Project Name: Ab Initio study of Thermoelectric Materials
Project Shortname: MATS0712
Discipline Name: Material Science
The Computational and Theoretical Physics (CTheP) Group at the Department of Physics, Masinde Muliro University of Science and Technology, comprised of a team of researchers and students dedicated to the development and application of computational modeling to outstanding problems in materials science. In general, the group seeks a deeper understanding of quantum-mechanical descriptions of interacting electrons and nuclei in order to predict and design properties of materials and devices using density functional theory as a tool for the first-principles simulation. This kind of simulations generates volumes of data which cannot be handled by the common desktop computers. High-performance computing provides the computational environment that includes parallel processing, large memory, and storage of the big data. Therefore, Center for High-Performance Computing facility in Cape Town, South Africa, plays a vital role in providing the resources that aid the discovery of new materials and accelerates its application in the industry, an achievement that would have been in vain with desktop computers. Some specific projects for our group include characterization of already known thermoelectric materials. The information gained from this project is used as identifiers in search for new thermoelectric materials with the tailored figure-of-merit. Such materials are used to enhance the efficiency of thermoelectric generator power system for renewable energy. Such outcomes of our projects have an impact on the livelihood of the rural communities in Kenya, and generally, supports the Agenda Four of the Kenyan economic blueprint that was recently availed by the president of the Republic of Kenya.
Principal Investigator: Dr Michael Owen
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-11-23
Allocation End: 2022-05-24
Used Hours: 116846
Project Name: Industrial heat exchanger simulation and analysis
Project Shortname: MECH1210
Discipline Name: Computational Mechanics
The Thermofluids Division at Stellenbosch University's Department of Mechanical and Mechatronic Engineering has a long history of research in industrial heat exchangers and has conducted pioneering work in the field of dry cooling for large thermal power plants in particular.
Research in this field continues with a focus on understanding and improving the performance of these water conserving cooling systems under a variety of ambient conditions (including strong winds). Work of this nature benefits greatly from the CHPC which allows us to investigate an unprecedented number of scenarios in a short amount of time using computational fluid dynamics (CFD) and parallel computing.
Dry cooling is a key technology for South Africa's energy sector since thermal energy production (coal, nuclear, combined cycle gas and solar thermal) is typically a water intensive process which South Africa can ill afford considering the arid nature of our country. Ensuring robust operation of dry cooling systems can contribute significantly to the sustainability of South Africa's energy production.
Principal Investigator: Prof Serestina Viriri
Institution Name: University of KwaZulu-Natal
Active Member Count: 25
Allocation Start: 2021-11-23
Allocation End: 2022-05-24
Used Hours: 27757
Project Name: Deep Learning in Medical Image Analysis
Project Shortname: CSCI1229
Discipline Name: Computer Science
The CHPC is providing state-of-the-art High-Performance Computing which enables to conduct of current innovative research, especially in Artificial Intelligence. Powerful parallel computation processing capabilities are required in every research, especially in this era of Big Data. In our case, we have used this platform to conduct the life-saving type of research through Medical Imaging Analysis. Some positive results achieved are Accurate Skin Lesion Segmentation, Classification of Osteoarthritis Severity from Knee X-ray Images Using CNN, and even beyond medical analysis to remote sensing, etc.
Principal Investigator: Dr Collins Obuah
Institution Name: University of Ghana
Active Member Count: 3
Allocation Start: 2021-11-24
Allocation End: 2022-05-25
Used Hours: 83502
Project Name: Bioinorganic Chemistry
Project Shortname: CHEM1351
Discipline Name: Chemistry
This research is on finding new catalysts which can oxidized ethylene to alcohols or epoxides . This research has show that group VII B metal complexes which fine-tuned can be good catalysts which can catalyze oxidation reaction of ethylene to exclusively alcohols or epoxides.
We have used the CHPC facility for the past two years to investigate the best group VII B metals for the oxidization of ethylene. All the research work were computational and the CHPC has been the only resource we have used to achieve some good success in our research.
Principal Investigator: Prof Rajshekhar Karpoormath
Institution Name: University of KwaZulu-Natal
Active Member Count: 14
Allocation Start: 2021-11-24
Allocation End: 2022-05-25
Used Hours: 55610
Project Name: Design of small novel organic compounds as potential drug candidates
Project Shortname: HEAL0835
Discipline Name: Health Sciences
Novel chemotherapeutic agents against multidrug resistant-tuberculosis (MDR-TB) are urgently needed at this juncture to save the life of TB-infected patients. In this work, we have synthesized and characterized novel isatin hydrazones 4(a-o) and their thiomorpholine tethered analogues 5(a-o). All the synthesized compounds were initially screened for their anti-mycobacterial activity against the H37Rv strain of Mycobacterium tuberculosis (MTB) under level-I testing. Remarkably, five compounds 4f, 4h, 4n, 5f and 5m (IC50 = 1.9 µM to 9.8 µM) were found to be most active, with 4f (IC50 = 1.9 µM) indicating the highest inhibition of H37Rv. These compounds were further evaluated at level-II testing against the five drug-resistant strains such as isoniazid-resistant strains (INHR1 and INH-R2), rifampicin-resistant strains (RIF-R1 and RIF-R2) and fluoroquinolone-resistant strain (FQ-R1) of MTB. Interestingly, 4f and 5f emerged as the most potent compounds with IC50 of 3.6 µM and 1.9 µM against RIFR1 MTB strain, followed by INH-R1 MTB strain with IC50 of 3.5 µM and 3.4 µM, respectively. Against FQ-R1 MTB strain, the lead compounds 4f and 5f displayed excellent inhibition at IC50 5.9 µM and 4.9 µM, respectively indicating broad spectrum of activity. Further, molecular docking, ADME pharmacokinetic and molecular dynamics simulations of the compounds were performed against the DNA gyrase B and obtained encouraging results.
Principal Investigator: Dr Gebremedhn Gebreyesus Hagoss
Institution Name: University of Ghana
Active Member Count: 4
Allocation Start: 2021-11-25
Allocation End: 2022-06-30
Used Hours: 287813
Project Name: Atomistic simulations and condensed matter
Project Shortname: MATS1031
Discipline Name: Physics
My research group at the Department of Physics, University of Ghana is focused on the study of the structural, electronic, magnetic, and optical properties of 2D and bulk materials. Computational simulations of the properties of these materials are studied using density functional theory and extended Hubbard functionals as implemented in QUANTUM ESPRESSO, GW approximation of many-body perturbation theory (MBPT), and the Bethe-Salpeter equation (BSE). Our computational work would not be possible without the enhanced computing resources available at the CHPC in South Africa. Typically, we prepare our input files on our laptops and works stations and submit the job to the CHPC cluster. The jobs are monitored at least once a day. The results are downloaded when completed. Our main areas of research: 1. Electronic, magnetic, structural, and optical properties of transition-metal oxides. First principle calculations were carried out to study the structural, electronic, optical, and magnetic properties of transition metal oxides. The results are compared to our experimental results. 2. Perovskite Materials for energy applications The accurate description of the structural, Electronic, and optical properties of low-temperature phase perovskite are not yet well established. In this project, we focus on the investigation of structural, electronic, dielectric, and vibrational properties of low-temperature phase perovskite materials using density functional theory with the inclusion of the onsite and inter-site Hubbard correction (DFT+U+V). 3. 2D Materials Research mostly transition metal dichalcogenides are compounds that we are interested to study in this project, have a chemical formula MX2 (where M = Ti, Mo, Nb, W, Hf, Re, etc. and X = S, Se, and Te). In particular, we focus on monolayers, bilayers, and heterostructure of transition metal disulphides (TMS2), as photo-catalysts for hydrogen evolution reaction. We are also continuing to study the excitonic properties of 2D HfS2 monolayer via lanthanide substitutional doping using GW and BSE. 4. Ruddlesden-Popper perovskite ruthenates In this project, we focus on calculating the electron-phonon couplings of Sr4Ru3O10 using DFT with the inclusion of the onsite Hubbard correction (DFT+U) and spin-orbit coupling.
Principal Investigator: Prof Martin van Eldik
Institution Name: North-West University
Active Member Count: 4
Allocation Start: 2021-11-29
Allocation End: 2022-06-30
Used Hours: 64190
Project Name: Thermal-Fluid modelling
Project Shortname: MECH1247
Discipline Name: Computational Mechanics
The Thermal-Fluid Modelling research group, is situated in the Faculty of Engineering at the North-West University. Students using the CHPC currently consists of 1 Masters student and 1 PhD. All students are using Star-CCM+ for their studies. The research group produced one masters student in 2020 that investigated the thermal fluid modelling of a titanium helically coiled fluted tube. The current Masters student is conducting an aeroacoustic analysis of a glider plane component and how to identify the source and replicate it. The PhD student is investigating the CFD (computational fluid dynamics) modelling of a sinter mixing drum to improve the production throughput. The CHPC makes it possible for the students to be able to complete their work on time, which otherwise would not have been able to.
Principal Investigator: Dr Richard Klein
Institution Name: University of the Witwatersrand
Active Member Count: 13
Allocation Start: 2021-11-29
Allocation End: 2022-05-30
Used Hours: 139368
Project Name: PRIME
Project Shortname: CSCI1340
Discipline Name: Computer Science
PRIME is a research lab in the School of Computer Science and Applied Mathematics at Wits University. We focus on Machine Learning and Artificial Intelligence techniques for Perception and Representation Learning. The primary goal of this research is to be able to learn good representations of data that is unlabelled. This means that in the future, AI can learn from large unlabelled datasets which are cheaper to create, and then fine-tune their understanding of the world on smaller, more expensive, labelled datasets. We have recently proven the robustness of self-supervised deep learning techniques to viewpoint changes, showing that models trained in this way not only require less labelled data, but that you get smarter, more robust predictive models at the end of the process. Other work has focused on the application of these deep learning techniques to microcrack detection in PV solar panels, helping to evaluate performance of solar modules over their lifespan.
Principal Investigator: Dr Steve Peterson
Institution Name: University of Cape Town
Active Member Count: 4
Allocation Start: 2021-11-19
Allocation End: 2022-05-20
Used Hours: 122646
Project Name: Development of a Prompt Gamma Imaging Device
Project Shortname: PHYS0830
Discipline Name: Physics
The UCT Prompt Gamma Imaging (PGI) Group is a part of the Measure Research Unit based in the Department of Physics at the University of Cape Town. PGI is a new form of imaging that will be used in conjunction with proton radiation therapy cancer treatments. The protons that are used to attack the cancer cells produce gamma-rays that we are detecting and using to produce images of the dose deposited in the patient. PGI is still a technology in development and relies heavily on computing resources for both detector design and image reconstruction.
Principal Investigator: Prof Andrei Kolesnikov
Institution Name: Tshwane University of Technology
Active Member Count: 3
Allocation Start: 2021-11-30
Allocation End: 2022-05-31
Used Hours: 90644
Project Name: Modeling and simulation of multiphase flows with application in processing industries
Project Shortname: MECH1386
Discipline Name: Chemical Engineering
Multiphase flows modeling research is conducted at Tshwane University of Technology.
A group of 3 M.Eng. and 2 D.Eng. students developing mathematical models and perform simulations of heat and mass transfer in plasma flows, and other high-temperature chemical and mineral processes. Recently investigation of corrosion mechanisms and development of new corrosion inhibitors also became of the group's interests.
The results of CFD modeling and simulations, molecular dynamics, quantum chemical simulations will allow to improve technological processes, save energy and reduce environmental risks.
Principal Investigator: Prof Ken Craig
Institution Name: University of Pretoria
Active Member Count: 4
Allocation Start: 2021-11-30
Allocation End: 2022-05-31
Used Hours: 332171
Project Name: CFD modelling of jet impingement with phase change
Project Shortname: MECH1296
Discipline Name: Computational Mechanics
Jet impingement heat transfer with boiling is a promising new heat transfer method for use in power electronics cooling. Most of the current research on this topic in literature is experimental, leaving a gap for computational studies to further investigate the relevant influencing parameters. Jet impingement flow is difficult to simulate because of its transient nature and multiple flow regimes involved (from laminar to turbulent). Previous research by Prof Ken Craig's group has developed techniques based on Large Eddy Simulation to evaluate the effect of round jets with and without swirl on solar receiver heat transfer surfaces. This program extends that work to having phase change as well in the form of boiling. The latent heat absorbed by boiling works in addition to the enhance convection provided by the jet. Various case studies are being developed for application in the electronics cooling industry and the approach has been validated using test cases from literature.
Principal Investigator: Dr Benjamin Victor Odari Ombwayo
Institution Name: Masinde Muliro University of Science and Technology, Kakamega, Kenya
Active Member Count: 5
Allocation Start: 2021-11-30
Allocation End: 2022-05-31
Used Hours: 7141
Project Name: Ab initio study of Solar Energy Materials
Project Shortname: MATS1426
Discipline Name: Material Science
Solar energy materials have been widely characterized experimentally using different techniques for a variety of properties such as structural, electrical, optical and defects. In search of suitable materials that can replace silicon in fabrication of efficient and stable solar cells, Inorganic–organic metal halide perovskite solar cells (PSC) are currently in the limelight of solar cell research due to their rapid growth in efficiency which has crossed 25% for laboratory scale devices but have stability and reliability issues. Our group, Computational and Theoretical Physics (CTheP) research group in Masinde Muliro University of Science and Technology, through the programme: Ab initio study of Solar Energy Materials (MATS1426), endeavor to investigate stability issues in perovskite solar cells and suggest remedies from a Density Functional Theory perspective. Our studies are on Triple Cation Mixed Halide perovskite materials of the form Cs0.5FA0.3MA0.2Pb(I0.2Br0.8)3 and are effectively calculated in a HPC environment due to the large number of atoms involved. We are investigating the intrinsic stability of the materials which constitute the band structure, defects, thermodynamic and phase stability. Knowledge gained will greatly help in adopting strategies to further improve the efficiency and stability of PSCs. This work is being done using the Quantum Espresso code in CHPC.
Principal Investigator: Prof Richard Tia
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 41
Allocation Start: 2021-12-03
Allocation End: 2022-06-03
Used Hours: 1842577
Project Name: Computational Chemistry and Materials Science
Project Shortname: CHEM1044
Discipline Name: Chemistry
The work being conducted in this research programme involves the development of active and selective catalysts for conversion of greenhouse gases like carbon dioxide into useful fuels and fine chemicals, the depolymerization of lignin, an abundant resource in plant biomass, into value-added chemicals and fuels and studies on the computer-aided synthesis of organic molecules of applications in medicine.
With increasing interest of the scientific community and government agencies on the mitigation of rising carbon dioxide levels and the attractiveness of using carbon dioxide as a feed-stock for industrial processes, this work has direct benefit for humanity.
Our research activities fall under two broad areas, namely;
1. Molecular modeling - Exploring the mechanisms of organic, inorganic and organometallic reactions.
These studies afford important molecular-level mechanistic details of industrially-relevant chemical reactions that are difficult, if not impossible, to obtain experimentally. Molecular level understanding enable chemical reactions to be guided quickly and efficiently along desirable pathways, with the effect of producing the right product with minimal energy consumption and minimal environmental impact. A key aspect of this work is the development of homogeneous transition metal catalysts for the depolymerization of lignin, an abundant natural resource, for fuel.
2. Materials modeling
The atomic- and molecular-level mechanistic insight gained from these studies are useful for the design of novel catalysts for the conversion of CO2 and other green-house gases as well as stranded gases from the petroleum refining into useful fuels. This addresses the problem of global warming as well as providing an environmentally friendly alternative sources of fuel.
The research objectives described here are very computationally-demanding and using the CHPC resources is highly beneficial. Without that, we would not be able to carry out some of these calculations.
Principal Investigator: Prof Bettine van Vuuren
Institution Name: University of Johannesburg
Active Member Count: 5
Allocation Start: 2021-12-06
Allocation End: 2022-06-06
Used Hours: 114528
Project Name: Biocomplexity on sub-Antarctic islands
Project Shortname: CBBI1153
Discipline Name: Bioinformatics
We are a research group at the Centre of Ecological Genomics and Wildlife Conservation based at the University of Johannesburg, South Africa.
Biodiversity is being lost at unprecedented rates. Factors such as climate change, habitat fragmentation, and environmental degradation (including alien species) are influencing the distribution and abundance of species, often in ways that are impossible to predict. As conservation geneticists, we are interested in exploring spatial and temporal genetic trends in a variety of organisms (plants, invertebrates, microorganisms) on sub-Antarctic islands, with a special focus on Marion Island.
Our aim is to investigate genetic patterns and structure in the context of environmental changes (e.g. climate change), with the intention of using our study species as proxies for monitoring global change and its impact on biodiversity as a whole.
To do this, we use multiple workflows to create next-generation sequencing (NGS) data. By using the CHPC cluster, we can perform phylogenetic and population genetic analysis on our NGS data. Since no other platforms can manage the size of the data indicated above, we would not be able to conduct our research without the CHPC facilities. We are pleased with our success thus far and are grateful to the CHPC for their assistance.
Principal Investigator: Mr Reshendren Naidoo
Institution Name: University of the Witwatersrand
Active Member Count: 3
Allocation Start: 2021-12-07
Allocation End: 2022-06-07
Used Hours: 8214
Project Name: Influence of Gas Phase Kinetics on a Pulverised Coal Flame
Project Shortname: MECH1298
Discipline Name: Computational Mechanics
The combustion of coal particles is a complex process and there are several physical processes at play. These processes are in turn highly dependent on the quality of coal being combusted and hence there must be direct link between the numerical methods employed and the true heterogeneous nature of coal. One of these facet is the way combustible material is released from coal particles. The chemical mechanisms that are used to predict the flame characteristics are important to understand the ignition/stability and emissions profile of a given combustion regime. Hence it is important to study these chemical mechanisms.
Principal Investigator: Prof George Amolo
Institution Name: Technical University of Kenya, Nairobi, Kenya
Active Member Count: 12
Allocation Start: 2021-12-07
Allocation End: 2022-06-15
Used Hours: 1564452
Project Name: Properties of Materials for Green Energy Harnessing
Project Shortname: MATS862
Discipline Name: Material Science
MATS862 is a Kenyan group located at the School of Physics and Earth Sciences, The Technical University of Kenya, Nairobi. We have for more than 10 years focused on the properties of hard materials and materials for energy conversion. Materials around us now influence the state of the environment, the kind of energies that we use as well as our own well being. The work we do is crucial as it is complementary to that being done by groups or industry carrying out experimental work at the research or even production stage. The complementing work is done using fundamental science and employs computing resources at the CHPC to simulate conditions approaching actual cases or even predictive and hence able to guide decisions on production. Actual calculations are based on the use of properties of atoms, electrons and molecules without any other assumptions to determine their most stable energy state using density functional theory approach. In other cases the determination of the most stable geometry is obtained by employing a molecular dynamics technique.
We are doing very well and are pleased by the continued support from the CHPC on the provision of computational resources.
Our group will host visitors from Africa (Central Africa) and Europe (Sweden) this year and next year, respectively, and continues to be involved in activities run by the Kenya Education Network (KENET) at the national level
Principal Investigator: Prof Orde Munro
Institution Name: University of the Witwatersrand
Active Member Count: 7
Allocation Start: 2021-12-13
Allocation End: 2022-06-13
Used Hours: 64570
Project Name: Bioinorganic Chemistry
Project Shortname: CHEM1065
Discipline Name: Chemistry
BIOINORGANIC RESEARCH GROUP – WITS UNIVERSITY
The Bioinorganic Research Group at WITS University is led by Prof. Orde Munro, who is currently the SARChI Chair of Bioinorganic Chemistry. At present, the WITS group comprises three post-doctoral research fellows (Drs. Catherine Slabber, Hermina De Wit, Zeynab Fakhar, and Stefan Coetzee), four PhD students, and one MSc student.
NATURE OF OUR WORK
The work mainly involves the design, synthesis, and testing of novel metallodrug candidates for chemotherapy. The compounds incorporate ligands that can facilitate and control the interaction of the metal ion with biomolecules such as DNA and/or proteins. Molecular simulations are used not only to design compounds with desirable biological behavior, but also to gain fundamental insights on their mechanisms of action and electronic structures. We are particularly interested in exploiting and controlling the reactivity of both endogenous and non-natural metal ions to kill drug-resistant pathogenic bacteria such as TB (Mycobacteria) and cancer cells.
WHY WE USE SIMULATIONS
Modern drug discovery uses in silico methods for a number of reasons, including screening large libraries of compounds for binding efficacy to enzyme targets. The enzyme targets are crucial proteins for the survival and replication of cancer cells, bacteria and viruses. Screening many thousands of compounds using computer programs is an efficient way to select a handful of compounds that can ostensibly bind to the enzyme target of interest because they have the correct shape and functional groups. The small library so generated may then be synthesized in the laboratory and tested against the pathogen of interest and the putative enzyme target. In this way, computational methods shorten the time required to find promising new compounds for further testing and development as the medicines of tomorrow. Finally, computer simulations offer important three dimensional details about how drug-like compounds inhibit their enzyme targets, thereby guiding optimization of the compound's structure for enhanced target affinity and, in principle, improved in vivo efficacy.
PROJECT PROGRESS
Over the last 4 years the project has produced 7 published papers in ISI-rated international journals and 1 US patent. Another 6 papers are in advanced draft form. 1 PhD, 2 MSc, and > 5 Honours students have graduated. Importantly, use of the CHPC has expanded to all students in the group from MSc to post-doctoral level. Regarding the overall thrust of the work, the CHPC resources have supported efforts to discover novel soluble epoxide hydrolase inhibitors, delineate the mechanism of action of gold(III) quinoline-amide topoisomerase II inhibitors, and design a library of novel metallodrug candidates for SARS-CoV-2. The latter compounds have been synthesized for in vitro screening against the virus and the project is being written up.
Principal Investigator: Prof Gerrit Basson
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2021-12-14
Allocation End: 2022-06-14
Used Hours: 9755
Project Name: CFD modelling of Hydraulic Structures in Rivers
Project Shortname: MECH0985
Discipline Name: Computational Mechanics
The Institute for Water & Environmental Engineering Stellenbosch University has 1 Doctoral students and 1 Post-Doctoral student who are investigating CFD modelling of hydraulic structures with complex vortices and sedimentation. Key aspects of the research are related to the improved prediction of bridge pier scour, to investigate the addition of flares on spillways, to reduce rock bed scour in plunge pools at spillways, to optimize the design of vortex settling basins, sand traps and settlers for sediment removal at hydropower and water abstraction plants. The simulations are too computationally intensive for standard computers and the CHPC assists in the research by allowing the students to more extensively explore the hydraulic behaviour of these structures.
Principal Investigator: Prof Rasheed Adeleke
Institution Name: North-West University
Active Member Count: 13
Allocation Start: 2021-07-28
Allocation End: 2022-01-26
Used Hours: 1656
Project Name: Environmental and Agricultural Microbiology
Project Shortname: CBBI1183
Discipline Name: Environmental Sciences
Our research programme at the Unit for Environmental Sciences and Management, North-West University, Potchefstroom continues to advance the understanding of microbial communities across different environments and agroecosystems such as plants, food, water and soils. Our core interest is towards soil and plant health. We are actively investigating the microbial endophytes associated with ready-to-eat vegetables with the aim of identifying novel and efficient endophytes and their roles in improving plant and human health. We seek to generate novel solutions that address critical global issues, including environmental sustainability, green energy and food security. Currently, our research projects utilize state of the art sequencing technologies to decipher the whole genome of novel microorganisms, and to understand the composition and functional repertoire of microbial community in different environments such as, food, soil, water and plant. Such data can only be analysed using large computer hardware/resources. Thus, we are very reliant on the CHPC's server to execute most of our bioinformatics analysis and/or pipeline. Our research programme currently boasts of several postgraduate students at honours, masters, PhD and postdoc level level.
Principal Investigator: Dr Phillip Nyawere
Institution Name: Kabarak University, Nakuru, Kenya
Active Member Count: 9
Allocation Start: 2022-01-03
Allocation End: 2022-09-21
Used Hours: 101351
Project Name: Barium Floride and Perovskites
Project Shortname: MATS0996
Discipline Name: Physics
Barium Fluoride and Perovskites is a group of researchers based in Kenya and collaborating from different universities both private and public led by Dr. Phillip Nyawere of Kabarak University. At Kabarak University we have a growing number of MSc students in Physics and our PhD program is starting soon. Our department of Physical and Biological Sciences is under school of Science, Engineering and Technology. We have a team of Professors and Senior Lecturers specialized in various fields of research.
When it was started, it was the Principal Investigator that had knowledge on the use of the CHPC and the computational code necessary for their research. Our simulation codes are Quantum Espresso and Siesta Code.
This has changed with time and graduates have been mentored through this facility. We have graduated one PhD and one MSc students. This year 2022 we expect two more to graduate with PhD and MSc.
Our team capacity has grown with the newly graduated members now able to also supervise the MSc team of students. I believe without this facility it would be difficult to carry our research at the comfort of our institutions and homes.
We research on materials from the simple structures to smart materials including the perovskites of organic-inorganic materials for photovoltaic applications.
We have also ventured into studying superconducting materials of perovskite nature. These are materials of zero resistance at high temperature which will help in energy conservation and is an area of great interest from both experimentalists and computational teams.
We therefore look forward to continuous collaboration between CHPC and our group for greater outputs in our field of research.
We are very grateful to the generous availability of the computational resources here with.
Principal Investigator: Prof Josua Meyer
Institution Name: University of Pretoria
Active Member Count: 6
Allocation Start: 2021-09-25
Allocation End: 2022-04-28
Used Hours: 633443
Project Name: Fundamentals of forced and mixed convection in heat exchangers
Project Shortname: MECH1094
Discipline Name: Computational Mechanics
The looming smart world of new generation, are more oriented towards numeric' s, rather than hardwired. The use of CHPC is now concluding many of our research, with a proper physical evidence. Managing big data's, iterating equations for its convergence, with supporting software of graphics and animation, has really became easier. The beauty of a research, physical justification and physics for flow and heat interaction has really become easier to define and write. Numeri's with high computational power is the new face of upcoming world of application in Aerospace/ Fluid and thermal engineering. The use of CHPC, has helped us in observing the molecular physics behind it.
Principal Investigator: Dr Gurthwin Bosman
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-07-27
Allocation End: 2022-01-25
Used Hours: 56708
Project Name: Quantum computations for photoinduced reactions
Project Shortname: CHEM1126
Discipline Name: Physics
The research group is one of the arms of the Laser Research Institute in the Department of Physics of Stellenbosch University. Laser Research Institute of the Stellenbosch University, boasts of ongoing formidable research projects in the fields of ultrafast science, spectroscopy and laser diagnostics, laser development, trapped ion quantum control, and biophotonics and imaging, with a number of laboratories housing ultra-modern equipment. Our research group is focused on ultrafast transient absorption spectroscopy on chemical compounds ranging from simple molecules to polymers. Experimental setups for ultrafast transient absorption spectroscopy have been thoroughly developed over the years by different masters, doctoral and postdoctoral students. The research projects conducted in the ultrafast laboratory had produced masters' and doctoral theses, as well as peer-reviewed publications.
Scientific observations being made by experiments in our research group need to be properly understood and interpreted for proper communication to the general community. Therefore, a level of theoretical simulation which will allow deeper insights is needed. Part of this theoretical simulation is the density functional calculation, a theory well accepted to interpret molecular and/or structural dynamics we observe experimentally. The calculations can take forever if run on a single computer. This is where the Centre for High-Performance Computing (CHPC) comes in.
Centre for High-Performance Computing (CHPC) offers a wide range of theoretical coding and simulation with the advantage of supercomputing on parallel computers. This enables users' jobs submission to run on a faster scale. Our research group is privileged to have access to these supercomputing capabilities and the provision of simulation software for our calculations.
We are running a series of density functional calculations ranging from simple molecules to polymers using Gaussian 09 that is provided by CHPC.
Principal Investigator: Dr Fabio Cinti
Institution Name: NITHEP
Active Member Count: 4
Allocation Start: 2021-07-27
Allocation End: 2022-01-25
Used Hours: 17410
Project Name: Quantum Monte Carlo for ultra-cold atoms
Project Shortname: PHYS0892
Discipline Name: Physics
A group led by Fabio Cinti (associate professor at University of Johannesburg) explores pattern formation of special structures like stripe phases, cluster crystals and quasicrystals in a quantum regime. Presently these structures can understand a large amount of fascinating phenomena in soft matter, superconductivity, nonlinear optical systems, .and long-range interacting systems in general. In this context quasicrystals are one of the most intriguing examples, as particles self-assemble in a long-range ordered pattern which is at the same time non-periodic, thus been able to exhibit forbidden crystalline ordering. Many studies observed cluster quasicrystals in soft macromolecular systems at finite temperatures by using this type of interactions. While a recent theoretical work have surprisingly revealed the stability of those structures also at zero temperature for a particular case, the extent to which classical cluster quasicrystals can be stable in the absence of thermal fluctuations is a matter of debate.
We investigate quantum cluster quasicrystals by imposing external quasi-periodic potentials to bosonic systems, so creating quasicrystalline structures in two-dimensional optical lattices. Interestingly, the competition of interactions and quasiperiodicity generate a wide range of phases, such as supersolidity and Bose glasses. We also observe superfluidity in a model relevant to quantum cluster quasicrystal. By using quantum Monte Carlo approaches, it was found that moderate quantum fluctuations make dodecagonal structures to persist, leading to a small but finite local superfluid phase.
By increasing fluctuations, a structural transition from quasicrystal to cluster triangular crystal takes place. Our studies bring to the conclusion that, at a quantum level, quasicrystal phases are produced as a joined effect of quantum fluctuations and a properly designed interaction potential between particles. This research have been carried out using the computational resources of the CHPC facilities.
Principal Investigator: Dr Moses Okpeku
Institution Name: University of KwaZulu-Natal
Active Member Count: 7
Allocation Start: 2022-01-11
Allocation End: 2022-06-30
Used Hours: 1672
Project Name: Genomics and Bioinformatics Group Westville Campus
Project Shortname: CBBI1185
Discipline Name: Bioinformatics
The Genomics and Bioinformatics group is resident in the Discipline of Genetics, School of Life Sciences, University of Kwazulu-Natal Westville. Our research focuses is on the genetic basis of infectious diseases and characterization studies. Current work on Malaria and COVID-19 gene-based vaccines is progressing with impressive results. We recently began characterization for endangered Sea Cucumber Spp. in South Africa. and depend heavily on the CHPC for most of the heavy lifting data analyses. My CHPC allocation is very important in validating what has been done and what we are doing now.
Principal Investigator: Prof Mario Santos
Institution Name: University of Western Cape
Active Member Count: 16
Allocation Start: 2022-01-11
Allocation End: 2022-07-12
Used Hours: 5899582
Project Name: Cosmology with Radio Telescopes
Project Shortname: ASTR0945
Discipline Name: Astrophysics
The Centre for Radio Cosmology (CRC) at UWC aims to fully explore the use of the next generation of radio telescopes for measurements in cosmology, in particular with MeerKAT, HERA and the SKA. Our work at the CHPC is mostly focused on running simulations of signals we expect to observe with MeerKAT/HERA/SKA and test how well radio telescopes can constrain cosmological models. The CHPC facilities already help to set cosmological constraints on observations made by MeerKAT and the HERA radio telescopes, both set in South Africa.
Principal Investigator: Dr ADEDAPO ADEYINKA
Institution Name: University of Johannesburg
Active Member Count: 8
Allocation Start: 2022-01-11
Allocation End: 2022-07-11
Used Hours: 898488
Project Name: Computational Design of Molecules
Project Shortname: CHEM1221
Discipline Name: Chemistry
The Computational Design of Molecules research group is based at the University of Johannesburg. We aim to design new and improved molecules and materials for applications in Catalysis, drug discovery and renewable energy. Since the availability of energy and drugs for neglected diseases are some of the main challenges faced by the populace on the African continent, being able to achieve our aims as a group will go a long way to facilitate the achievement of sustainable development goals on the continent
Principal Investigator: Prof Juliet Pulliam
Institution Name: Stellenbosch University
Active Member Count: 20
Allocation Start: 2022-01-11
Allocation End: 2022-07-19
Used Hours: 53959
Project Name: South African Centre for Epidemiological Modelling and Analysis
Project Shortname: CBBI1106
Discipline Name: Other
SACEMA is a DSI-NRF Centre of Excellence focused on Epidemiological Modelling and Analysis, hosted at Stellenbosch University. We use computational approaches to address questions of public health relevance to South Africa and the African continent. We use the CHPC primarily to run stochastic simulations and for simulation-based validation of novel inference methods.
Principal Investigator: Prof Penny Govender
Institution Name: University of Johannesburg
Active Member Count: 6
Allocation Start: 2022-01-11
Allocation End: 2022-07-12
Used Hours: 12812
Project Name: Design of MOFs-based sorbent materials for capturing carbon dioxide
Project Shortname: CHEM0797
Discipline Name: Material Science
Environmental and energy problems are important topics globally due to the fast development of urbanization, huge population increases and industrialization. Recently, water pollution is a major issue in developing countries, particularly in regions where people rely on groundwater for domestic and drinking purposes. Almost 40 % of the world's population is facing water shortage. Most countries depend on the low-efficient power generation from the burning fossil fuels as their source of energy. The burning of fossil fuels normally generates a huge amount of CO2, which induces environmental concerns, such as greenhouse effect. As non-renewable natural resources, fossil fuels cannot be sustainable and depleted shortly in the next century owing to their environmental issues and high consumption rate. This has led to the investigation of photocatalysis as a possible solution to the problem. Despite significant concern and substantial efforts, the design of semiconductor photocatalyst materials that is both efficient and robust enough for water/wastewater treatment and energy production is still lacking. Due to the complexity of the experimental process, it is difficult to know the mechanism and origin of the enhanced photocatalytic performance. The first-principles approach is highly desirable in the area of materials science, particularly for understanding both the photocatalysis of nanomaterials, as well as the subsequent physical and chemical properties. Density functional theory (DFT) calculations was used as the method of clarification as it has already proven its superiority in studying the structural stabilities, electronic properties, work functions, charge transfer, and carrier effective mass of several semiconductor photocatalyst materials due to its reduced computational cost in calculating the electron density rather than the wave function.
Principal Investigator: Prof Mahmoud soliman
Institution Name: University of KwaZulu-Natal
Active Member Count: 35
Allocation Start: 2022-01-12
Allocation End: 2022-07-13
Used Hours: 1929132
Project Name: Drug Design, development and modelling
Project Shortname: HEAL0790
Discipline Name: Health Sciences
The research scope of the Molecular Bio-Computation and Drug Design Research Lab covers a wide range of computational and molecular modeling research areas with main focus on biological systems and drug design approaches. Main interest is related to design and study of biologically and therapeutically oriented targets by employing the applications of computational methods to the study of problems of chemical and biochemical reactivity, with particular focus upon the transition state, environmental effects on mechanisms, the origins of catalysis, and the interpretation of kinetic isotope effects. This includes mechanistic pathways and transition states for reactions in enzyme and solutions; design of enzyme inhibitors and exploring the binding and catalytic theme of the designed targets and adopting sophisticated computational approaches to understand protein structures and functions. We show keen interest in diseases of global burden and evident prevalence within the south African populace such as HIV/AIDS, Tuberculosis and Cancer. The computational resources provided by CHPC is employed in performing the molecular dynamic calculations and the corresponding post molecular dynamic simulation analysis
Principal Investigator: Dr Holliness Nose
Institution Name: Technical University of Kenya, Nairobi, Kenya
Active Member Count: 2
Allocation Start: 2022-01-12
Allocation End: 2022-07-13
Used Hours: 198464
Project Name: Inorganic Computational Chemistry
Project Shortname: CHEM1003
Discipline Name: Chemistry
Polymerization of olefins catalyzed by metal catalysts has been a focus of studies in the last several decades because of its importance in industries such as synthetic polymers, detergents, plasticizers, lubricants among others. However, finding a catalyst that is specific, selective, and effective in production is quite challenging. In the search for new and more effective catalysts, our research is involved in the adjustment of the ligand and the metal in a catalyst in an effort to improve its catalytic activity and polymer properties. Mechanistic studies are also involved in order to unravel the complex catalyzed olefin polymerization reactions. The mechanisms involve the following elementary reactions: (a) Catalyst activation by co catalysts, (b) Polymer chain initiation, which is believed to proceed via the Cossee-Arlman mechanism, (c) Polymer chain propagation reaction that can yield either linear polyethylene or branched polyethylene, and (d) Polymer chain termination/transfer processes such as -H exchange between monomer and polymer and -H elimination processes. Understanding the mechanisms of these elementary reactions is extremely important and will allow us to search for new and more effective catalysts for olefin polymerization and control the structure of the polymers. It is against this backdrop that we intend to design the P^N hemi labile transition metal complexes and study their mechanisms theoretically. The designed templates will be forwarded to our collaborators for synthesis and a further study of their structure-activity relationships through reaction with ethylene monomers will be carried-out.
Principal Investigator: Dr Kevin Lobb
Institution Name: Rhodes University
Active Member Count: 14
Allocation Start: 2021-12-30
Allocation End: 2022-06-30
Used Hours: 1119154
Project Name: Computational Mechanistic Chemistry and Drug Discovery
Project Shortname: CHEM0802
Discipline Name: Chemistry
In the computational mechanistic chemistry and drug discovery group we have two complementary interests - how do reactions occur, and how can we inhibit disease progression. Sometimes we look at the reactions that take place in the inhibition or progress of disease and our interests overlap.
Recent studies have looked at strategies for generating transition states (using DFT) at a large scale - a pilot study generates 2000 TS's with few problems. In terms of disease we have recently looked at compounds that inhibit HIV protease, and the behaviour of the spike protein for COVID-19 under various temperatures.
Principal Investigator: Prof Johan W. Joubert
Institution Name: University of Pretoria
Active Member Count: 2
Allocation Start: 2022-01-05
Allocation End: 2022-07-06
Used Hours: 185550
Project Name: Mobility modelling using vehicle telematics
Project Shortname: MECH0803
Discipline Name: Other
Government (all spheres) is tasked with the challenge of deciding where to invest in infrastructure (for housing/transport/economic development etc). The real hard part is to balance two questions. Firstly, "who gets the benefit of infrastructure?" and, secondly, "who PAYS for those benefits?". In our research programme we develop agent-based transport models that are able to depict and imitate the complex, multi-modal and large-scale transport scenarios in a South African context. Each individual - from the Metrorail commuter, to the private car owner, to the minibus taxi driver, to the commercial vehicle delivering your goodies at the Woolies Foods - are represented in such a transport model. The goal of our ongoing research is to assist government and industry make better informed, evidence-based decisions about the intended, and unintended consequences of specific transport interventions. Since every individual is modelled the computational burden is substantial and the CHPC allows us to overcome the data preparation and simulation challenges of executing the large-scale, MATSim models.
Our research extends the scope of freight impact to the international level by looking at the vulnerability of maritime connectivity, especially as it relates to the effect that Covid has had on international freight movement - a key aspect to keep economies running.
Principal Investigator: Prof Willem van Otterlo
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2021-12-30
Allocation End: 2022-10-19
Used Hours: 9020
Project Name: Small molecule bioactives
Project Shortname: CHEM1326
Discipline Name: Chemistry
Our research group focuses on the design and synthesis of small bioactive molecules within the Group of Organic and Medicinal Organic Chemistry (GOMOC) in the Department of Chemistry and Polymer Science at Stellenbosch University. Our group utilizes the Schrodinger modelling package through the CHPC to design small molecule inhibitors of key proteins implicated in important disease states including cancer, rheumatoid arthritis and viral infections (like that from SARS-COV-2). All of these diseases have an important impact on the quality and overall productivity of human life, and it would therefore be valuable to have access to other small molecules as inspiration for novel therapeutics. The access to the state-of-the-art modelling Schrodinger package is critical to the design of small libraries of compounds that are then synthesized in our laboratories, followed by bio-evaluation of these molecules. Subsequent structure-activity relationships, of great value for the design of next-generation libraries, can then also be determined through this software.
Principal Investigator: Dr Sunita Kruger
Institution Name: University of Johannesburg
Active Member Count: 4
Allocation Start: 2022-01-14
Allocation End: 2022-07-15
Used Hours: 1462727
Project Name: Environmental Heat Transfer
Project Shortname: MECH0995
Discipline Name: Computational Mechanics
This research group in the Mechanical Engineering Science Department of the University of Johannesburg is currently focusing on environmental heat transfer and fluid-structure interaction. Specifically heat transfer in naturally ventilated greenhouses, and the modelling of rock drills. Greenhouses are used to protect plants from adverse weather conditions and insects. Ventilation of greenhouses are of vital importance to ensure quality crop production. If temperatures in a greenhouse is too high, poor plant growth may result, and an increased need for frequent watering. A mechanical ventilation system might be required to cool the inside of the greenhouse. Natural ventilation is an alternative option used to ventilate greenhouses. Natural ventilation uses temperature and wind to control the indoor climate of greenhouses. Unfortunately greenhouses are extremely energy intensive. Energy costs are the third highest cost related to greenhouse crop cultivation. Reducing the operating costs of energy associated with greenhouse cultivation may result in a price reduction of greenhouse cultivated crops. Conducting experimental work on ventilation of greenhouses can be costly and cumbersome. Using computational methods such as CFD (Computational Fluid Dynamics) to obtain qualitative and quantitative assessment of greenhouses can reduce costs and time involved. The computer cluster at the Centre for High Performance Computing has been used to conduct the numerical investigation using CFD. Currently research is being conducted on a large rooftop greenhouse. Smaller greenhouses containing a single and multi-spans have also been investigated. Currently, three dimensional models of greenhouses subject to buoyancy driven flow are being investigated, as well as the fluid-structure interaction to improve the efficiency of rock drills.
Principal Investigator: Mr Randall Paton
Institution Name: University of the Witwatersrand
Active Member Count: 8
Allocation Start: 2022-01-14
Allocation End: 2022-07-15
Used Hours: 170766
Project Name: Compressible Gas Dynamics
Project Shortname: MECH0847
Discipline Name: Other
The work group led by Dr Randall Paton is looking at the effects of rapid, arbitrary acceleration, and specialised geometry, on compressible aerodynamics continues. This work is also being extended to examine some effects in space propulsion, and isentropic interactions of compressible flows. There are also plans to implement machine learning techniques to expand computational modelling capabilities.
The CHPC is an integral part of this work. It allows for the parametric simulation of these flow fields, which are often impossible to produce experimentally. The current cohort of several Masters and two PhD students all make use of this incredible facility (expertly managed by its tirelessly-dedicated staff) to expand the boundaries of knowledge, leading graduates from the group of all ethnic backgrounds and sexes to careers both locally and abroad.
Principal Investigator: Dr Fabio Cinti
Institution Name: NITHEP
Active Member Count: 3
Allocation Start: 2022-01-14
Allocation End: 2022-07-15
Used Hours: 174360
Project Name: Quantum Monte Carlo for ultra-cold atoms
Project Shortname: PHYS0892
Discipline Name: Physics
The group led by Fabio Cinti (visiting associate professor at University of Johannesburg) investigates systems characterized by Bose glass phases via a system of particles confined in external potentials. The Bose glass is an insulating phase that retains local superfluidity. Such a phase cannot appear in periodic systems, but is expected to exist in the presence of disorder. Quasi-periodic potentials, which exhibit long-range order but are not translationally invariant, offer an alternative geometry capable of giving rise to regions of local superfluidity constrained, precisely, by long-range order. We address the question of characterizing a Bose glass phase in two-dimensional quasi-periodic systems by studying its properties at finite temperature. In the first published results, we observed that the Bose glass phase survives at finite temperaure between the superfluid and the insulating phase for intermediate values of the quasi-periodic potential. Interestingly, the parameter regime used turns out to be compatible with experimental setups. Characterization of quantum glass is done by introducing ``zonal estimators'' that give access to local information on superfluidity and compressibility in finite regions. Our work paves the way for further investigation of these systems, such as ensembles of interacting bosons via dipolar potentials in quasi-periodic geometries.
We employ quantum Monte Carlo methods, in particular the Path integral Monte Carlo that allows to give a classical description of quantum many-body systems at finite temperature. Each physical system under investigation requires the introduction of new functions and techniques. Physical systems of interest consist of large numbers of particles, with quantum nature adding an additional degree of computational complexity: algorithms require efficient programming and parallelization, and even then they can run for days, making the use of clusters a necessity. With its specifications, and high performance, CHPC is perfectly suited to our needs, and we rely heavily on it in our research.
Principal Investigator: Prof Thomas Scriba
Institution Name: University of Cape Town
Active Member Count: 2
Allocation Start: 2022-01-14
Allocation End: 2022-07-15
Used Hours: 109654
Project Name: Tuberculosis vaccine and biomarker development
Project Shortname: HEAL1390
Discipline Name: Health Sciences
The research programme – Tuberculosis Vaccine and Biomarker Development – is one of the research programmes within SATVI (the South African Tuberculosis Vaccine Initiative), a world leading TB research entity located within the Faculty of Health Sciences at the University of Cape Town. The research entity performs cutting edge clinical and immunological research in TB pathogenesis, biomarker development and clinical vaccine development.
One of the major obstacles in tackling TB disease is the poor understanding of the various stages in which the disease progresses in the body, and of particular interest the "subclinical" disease stage in which the individual exhibits no clinical symptoms but has radiological abnormalities or microbiological evidence of active TB disease. A deeper understanding this disease stage would make it possible to identify individuals who are in fact at this stage of the disease, although clinically asymptomatic, and treat them before they advance to the active clinically symptomatic disease state. Moreover, as some those in this subclinical stage are now known to transmit the infection, their identification and treatment would limit onward transmission of the disease.
In light of this, this particular research project aims to use transcriptomic profiling to understand subclinical and the other stages of TB disease progression in the body, and ultimately develop a diagnostic tool that can be used to tease apart individuals with subclinical TB disease from healthy ones.
The analysis of such a dataset, which is terabytes of data, is infeasible on standard laptop and desktop computers owing to memory and computing power requirements. Upon obtaining the raw data, we rely on the computing power and resources provided by CHPC to the analysis.
We are currently on course in achieving the objectives of this research project. The results of the study shall be made available to the general public in published articles.
Principal Investigator: Dr Leigh Johnson
Institution Name: University of Cape Town
Active Member Count: 2
Allocation Start: 2022-01-14
Allocation End: 2022-08-01
Used Hours: 18223
Project Name: MicroCOSM: Microsimulation for the Control of South African Morbidity and Mortality
Project Shortname: HEAL1049
Discipline Name: Health Sciences
The MicroCOSM project, based at the Centre for Infectious Disease Epidemiology and Research at the University of Cape Town, is a project to simulate the spread of infectious diseases, as well as the incidence of non-infectious diseases, in the South African population. By simulating the social and biological factors that contribute to disease transmission, we can better understand which sub-populations need to be targeted for special interventions. We can also evaluate the impact that various prevention and treatment programmes have had in South Africa to date, and evaluate the potential impact of new programmes. This simulation involves generating nationally representative samples of 20,000-40,000 South Africans and tracking them over their life course. Because this involves many individual-level calculations, the model requires substantial computing power, and the CHPC is therefore critical to the conduct of these simulations. We have recently used the model to assess strategies to reduce the future incidence of cervical cancer in South Africa, and to assess the impact of different HIV programmes on trends in HIV incidence.
Principal Investigator: Dr Madison Lasich
Institution Name: Mangosuthu University of Technology
Active Member Count: 8
Allocation Start: 2022-01-17
Allocation End: 2022-07-18
Used Hours: 329853
Project Name: Chem Thermo
Project Shortname: CHEM1028
Discipline Name: Chemical Engineering
The Thermodynamics, Materials and Separations Research Group (TMSRG) at Mangosuthu University of Technology focuses on modelling and analyzing improvements to industrial processes to improve energy efficiency and reduce the negative impacts of the chemical and allied industries. The high performance computing resources of the CHPC are invaluable for multi-scale modelling of industrial processes and for detailed molecular simulations used to understand the fundamental behaviour of different fluids, gases, and materials. Thus far the TMSRG has produced 9 peer-reviewed articles and 9 international conference papers using the resources of the CHPC.
Principal Investigator: Dr Krishna Govender
Institution Name: University of Johannesburg
Active Member Count: 10
Allocation Start: 2022-01-17
Allocation End: 2022-07-18
Used Hours: 392428
Project Name: Computational approaches to design novel anticancer agents
Project Shortname: CHEM0792
Discipline Name: Chemistry
The work conducted in this research programme is purely computational. As such access to resources such as that provided by the CHPC is imperative in order to ensure that my research can be done in a timely fashion.
There are typically large molecular datasets that we work with when trying to determine drug candidates and if these where to be run on local desktop computers or laptops it would take weeks to months to produce a result. This same process is done on the CHPC within a day or two.
We are starting to look into taking results produced from the simulations run at the CHPC into the lab in order to produce better drugs for the treatment of COVID19 and cancer for example.
Principal Investigator: Dr Ikechukwu Achilonu
Institution Name: University of the Witwatersrand
Active Member Count: 6
Allocation Start: 2022-01-17
Allocation End: 2022-08-31
Used Hours: 135496
Project Name: Computational approach to modelling druggable proteins of neglected tropical diseases parasites and ESKAPE pathogens
Project Shortname: CHEM0849
Discipline Name: Bioinformatics
Press Release
Neglected Tropical Disease and ESKAPE Drug Discovery Biology Team, School of Molecular and Cell Biology, University of the Witwatersrand, Braamfontein Johannesburg 2050.
Structural Biology and Biochemistry is a broad interwoven discipline that requires knowledge of both basic chemistry, biology and physics. The theme of our research is Computational and Empirical aspects of Drug discovery research towards the development of new generation antihelminthics and antimicrobials. Our work requires the complementation of empirical facts using theoretical/computational approaches. We simulate ligand-protein interaction using induced fit molecular docking, we simulate protein array technology using high throughput virtual screening of molecular libraries such as the DrugBank and the PubChem databases, we also simulate affinity and stability of Ligand-protein complexes using molecular dynamic simulation. All these algorithms can be extracted from the CHPC as well as some free open-source software.
At the moment, the software and computing platforms available to my research group at the CHPC has made our research highly relevant. It has become easier to publish results derived by integrating empirical and computational studies. Moreover, without the CHPC, it would have been extremely difficult to acquire some of these software such as the Schrodinger Molecular Modelling Suite, which is over a million rand for a year licence.
Principal Investigator: Dr Pritika Ramharack
Institution Name: Medical Research Council
Active Member Count: 9
Allocation Start: 2022-01-17
Allocation End: 2022-07-18
Used Hours: 389919
Project Name: Phytomedicine in Metabolic disorders
Project Shortname: HEAL1387
Discipline Name: Health Sciences
The aim of my new Molecular modeling and Bio-computation research group within the Biomedical Research and Innovation Platform (BRIP), SAMRC, is to conduct predictive biological target identification, compound physio-chemical descriptions, molecular modeling, molecular docking and molecular dynamic simulations that are required for the enhancement of current therapeutic regimens in various metabolic disease conditions. These techniques may also be implemented in identifying and optimizing vaccine developments against SARS-CoV-2 variants amidst the Covid-19 pandemic. The group is currently still being established, with students currently being recruited for the 2021/2022 academic cycles. The work will focus on the use of the Schrodinger suite for the design and characterization of newly synthesized compounds and co-crystals, as well as the use of Glide for molecular docking. The use of the Amber suite will also be utilized to simulate a theoretical experimental environment that will be programmed using specialized chemical forcefields, thus allowing for molecular interactions and free-binding energy of the complexes to be analysed. This will provide critical information on the potential structural mechanisms of action of the compounds, as well as the structural dynamics of enzymes, with particular interest on mutational modifications. To perform these studies, the use of the CHPC will be critical in accessing the Schrondinger and Amber suites and to perform large scale molecular simulations. The successful use of the CHPC in my research is documented in various studies that are evidenced in 22 internationally peer-reviewed Journal articles (https://orcid.org/0000-0001-5850-6782?lang=en). My goal is continue utilizing this platform to facilitate and expand computational chemistry capacity development within South Africa, focusing on previously disadvantaged universities. The collaborative work within the SAMRC, using CHPC, will also provide key genomic and proteomic insights on new and circulating SARS-CoV-2 variants detected in South Africa.
Principal Investigator: Dr Gurthwin Bosman
Institution Name: Stellenbosch University
Active Member Count: 3
Allocation Start: 2022-01-17
Allocation End: 2023-02-02
Used Hours: 38260
Project Name: Quantum computations for photoinduced reactions
Project Shortname: CHEM1126
Discipline Name: Physics
The research group is one of the arms of the Laser Research Institute in the Department of Physics of Stellenbosch University. Laser Research Institute of the Stellenbosch University, boasts of ongoing formidable research projects in the fields of ultrafast science, spectroscopy and laser diagnostics, laser development, trapped ion quantum control, and bio-photonics and imaging, with a number of laboratories housing ultra-modern equipment. Our research group is focused on ultrafast transient absorption spectroscopy on chemical compounds ranging from simple molecules to polymers. Experimental setups for ultrafast transient absorption spectroscopy have been thoroughly developed over the years by different masters, doctoral and postdoctoral students. The research projects conducted in the ultrafast laboratory produced masters' and doctoral theses, as well as peer-reviewed publications.
Scientific observations being made by experiments in our research group need to be properly understood and interpreted for proper communication to the general community. Therefore, a level of theoretical simulation which will allow deeper insights is needed. Part of this theoretical simulation is the density functional calculation, a theory well accepted to interpret molecular and/or structural dynamics we observe experimentally. The calculations can take forever if run on a single computer. This is where the Centre for High-Performance Computing (CHPC) comes in.
Centre for High-Performance Computing (CHPC) offers a wide range of theoretical coding and simulation with the advantage of supercomputing on parallel computers. This enables users' jobs submission to run on a faster scale. Our research group is privileged to have access to these supercomputing capabilities and the provision of simulation software for our calculations.
We are running a series of density functional calculations ranging from simple molecules to polymers using Gaussian 09, ORCA, and Maestro that are provided by CHPC.
Principal Investigator: Prof Fernando Albericio
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2022-01-18
Allocation End: 2022-07-19
Used Hours: 70218
Project Name: Peptide chemistry and Organic chemistry
Project Shortname: CHEM1102
Discipline Name: Chemistry
We are Peptide Science Group at UKZN Westville campus. The project was brought up with an intend to learn computational chemistry. The project involves synthesis of several organic molecules towards application in oligonucleotide chemistry. With the help of computational tools we aim to predict the reaction mechanism. Since CHPC provides the server which speeds up the calculations and helps in achieving the preliminary results. The project so far had been very promising in predicting the mechanism. We expect to be going ahead with these results for further analysis towards prediction of reaction mechanism.
Principal Investigator: Prof Eric van Steen
Institution Name: University of Cape Town
Active Member Count: 8
Allocation Start: 2022-01-18
Allocation End: 2022-07-19
Used Hours: 741811
Project Name: Lateral interactions on surfaces
Project Shortname: CHEM0780
Discipline Name: Chemistry
(for our methane work):
The production of chemicals from natural gas is a dream reaction for chemists. A group from the University of Cape Town has been successful in the formation of an important chemical intermediate by selectively oxidizing the rather unreactive methane over platinum-based catalysts with very high selectivity at commercially interesting activities. High performance computing played a crucial role in determining the window of operation for this reaction and in elucidating the reaction pathway.
Principal Investigator: Mr Muaaz Bhamjee
Institution Name: University of Johannesburg
Active Member Count: 12
Allocation Start: 2022-01-19
Allocation End: 2022-07-20
Used Hours: 100927
Project Name: Modelling of Multiphase Flow in Process Equipment
Project Shortname: MECH0837
Discipline Name: Computational Mechanics
The research group is from the Department of Mechanical Engineering Science at the University of Johannesburg. The primary area of interest in the research is the use of coupled computational fluid dynamics and computational granular dynamics models to investigate the complex multiphase interaction in process equipment.
The models are based on multiple continuum and meso-scopic based approaches which are couple with models such as the Discrete Element Method, Immersed Boundary Method and Eulerian Dense Discrete Phase Models. Both commercial and open-source software has been used in the research. Primarily the research has been focused on
hydrocyclones. However, the research is moving onto fluidised beds, spiral concentrators, highly concentrated slurry flows and gas cyclones. The models are benchmarked accuracy) against experimental measurements as well as for computational efficiency. Accurate and computationally efficient modelling of such phenomena is vital to the design of hydrocyclones which is crucial to the mining industry. The research group has also begun to use this platform to model fluid dynamics in biomedical applications. The research group has produced numerous publications and has supported the research of four post-graduate students. Research into the computational efficiency of the models is continuing with the aid of the Centre for High Performance Computing (CHPC) facilities. Due to the complex physics and fidelity required in the problems tackled by the research group large computational resources are required. The CHPC provides a computational platform that is vital to the ongoing work and success of the research group.
Principal Investigator: Prof Shahida Moosa
Institution Name: Stellenbosch University
Active Member Count: 2
Allocation Start: 2022-01-19
Allocation End: 2022-07-20
Used Hours: 75819
Project Name: Rare Disease Genomics
Project Shortname: HEAL1396
Discipline Name: Health Sciences
The Undiagnosed Disease Programme (UDP) in South Africa recently published a paper reporting on the first 100 exomes performed on patients and families with Rare Diseases. The diagnostic yield was 51%, meaning that half of enrolled families received a diagnosis through the programme. As southern Africa's only UDP, we have shown the utility and success of exome sequencing even in our understudied populations.
Principal Investigator: Dr David Pearton
Institution Name: Oceanographic Research Institute (ORI)
Active Member Count: 3
Allocation Start: 2022-01-19
Allocation End: 2022-08-17
Used Hours: 17670
Project Name: Responses of South African Coral Reefs to Climate Change
Project Shortname: CBBI1072
Discipline Name: Environmental Sciences
With the threats of climate change, development, overfishing and pollution we are losing biodiversity at an alarming rate. Many say that we already at the beginning of a new mass extinction event to rival the end of the dinosaurs. We are losing species faster than we can discover and describe them - an incalculable loss to science and human well-being. In order to protect our biodiversity we need to know what is there and this means we need new tools to study it before it is lost forever. At the Oceanographic Research Institute we are using new genomics techniques and the CHPC to analyse the cryptic biodiversity occurring on the reefs, in sediments and in the water in and between Marine Protected Areas along the East Coast of South Africa. The huge range of species from all branches of the tree of life ranging from single celled organisms to worms, crabs and fish are too diverse and numerous to be identified by conventional means so we have used Autonomous Reef Monitoring Structures combined with Next Generation DNA sequencing and big data bioinformatics to identify the myriad of creatures living in our precious and threatened marine ecosystems. We have identified thousands of species from all branches of the tree of life, many never before seen in South Africa and others completely new to science. This enhanced understanding of biodiversity will enhance our understanding of these critical ecosystems and how they function as well as give us clues on how best to protect them for suture generations.
Principal Investigator: Prof Charalampos (Haris) Skokos
Institution Name: University of Cape Town
Active Member Count: 8
Allocation Start: 2022-01-19
Allocation End: 2022-07-20
Used Hours: 1422506
Project Name: Chaotic behavior of Hamiltonian systems
Project Shortname: CSCI1007
Discipline Name: Applied and Computational Mathematics
Using modern numerical techniques of Nonlinear Dynamics and Chaos Theory we investigate in a unified mathematical way the behavior of models describing the energy transport in disordered and granular media, as well as the properties of new elastic materials like graphene and molecules like the DNA. This research is performed by members of the 'Nonlinear Dynamics and Chaos Group' in the Department of Mathematics and Applied Mathematics at the University of Cape Town. The performed investigations constitute an innovative combination of Applied Mathematics, Chaos Theory, Hamiltonian Dynamics and Nonlinear Lattice Dynamics. Their outcomes will provide answers to some fundamental questions about the effect of chaos on the energy transport in disordered and granular media, and on the behavior of graphene and DNA. Some of the questions we try to address are: Does the presence of impurities and nonlinearities enhance or suppress the propagation of energy (e.g. heat) in solids, of light in crystals, and of vibrations in granular material? How does the ratio of the different base pairs in DNA chains affect their structural stability and the temperature at which the double-stranded DNA breaks to single-stranded DNA?
Principal Investigator: Dr Didam Gwazah Adams Duniya
Institution Name: BIUST
Active Member Count: 1
Allocation Start: 2022-01-20
Allocation End: 2022-07-19
Used Hours: 374349
Project Name: Cosmological Probes of Gravity, Dark Energy, and Large Scale Structure
Project Shortname: ASTR1480
Discipline Name: Astrophysics
A key problem in modern cosmology is to understand the true nature of dark energy (the cosmic component believed to be responsible for the accelerated expansion of the Universe at late-times). To this day, there is still no definitive answer from observational analysis or fundamental theory as to the true form of dark energy. Nevertheless, in 2017 the first observational evidence for the existence of an interaction between dark energy and dark matter -- in the form of a reciprocal exchange of energy and momentum -- was presented. (Note that when dark energy interacts with dark matter in this manner, it is referred to as Interacting Dark Energy, IDE.) This provides a strong step forward towards correctly identifying the nature of dark energy. Despite this achievement, more work needs to be done. The signature of IDE needs to be correctly identified in order to avoid misinterpretation of the effect and hence nature of dark energy. Our paper, "A probe of interacting dark energy with redshift space distortions," investigates the effect of IDE in redshift space distortions, using two models of IDE.
It should be pointed out that, in order to perform a proper analysis of any IDE models, one requires to solve a complex system of coupled differential equations many times -- from hundreds of thousands to several millions of times. An ordinary PC would take very many months to accomplish that. Only a super/high performance computer, like the facilities at the CHPC, which have superior speed and handle thousands of simultaneous computations can accomplish this in reasonable time; yet providing results of high fidelity. The CHPC has kindly provided these computing facilities for our current paper; with several others in preparation.
Principal Investigator: Prof Marelie Davel
Institution Name: North-West University
Active Member Count: 10
Allocation Start: 2022-01-21
Allocation End: 2022-08-03
Used Hours: 243217
Project Name: DNN Analysis
Project Shortname: CSCI1191
Discipline Name: Computer Science
MUST Deep Learning is a research group within the Faculty of Engineering at NWU. We perform basic and applied research in machine learning, with an emphasis on the theory and application of deep learning. Our theoretical work focuses on generalisation in deep learning and the interpretability of deep learning models. Building on a strong track record in the application of machine learning to multilingual speech processing, our current application domains are diverse, ranging from speech processing to space weather prediction to industrial applications of deep learning.
In the past decade, the field of Deep Neural Networks (DNNs) has brought renewed energy and focus to AI, through a series of remarkable breakthroughs in fields as diverse as speech recognition, board games and self-driving cars. In these and other applications, DNN systems have reached previously unknown levels of accuracy, making human-level performance a distinct possibility and thus suggesting novel insights on the mind-matter problem.
The successes of DNN systems have inspired much research into better algorithms, novel applications and a better understanding of DNNs. The MuST group is involved in all these aspects of DNN research. For example, we use DNNs and word embeddings to develop better language models for under-resourced languages; these models can be used in tasks such as speech recognition and machine translation. We also use DNNs to handle poor quality audio in speech and speaker recognition systems better, probe the processes at play during solar flare eruptions, and even optimising the design process of airfoil shape with some of our industry partners. We furthermore work with the industry partners on traffic flow prediction and channel state estimation (deep learning with telecommunications). We balance these applications with theoretical work focused on understanding and characterising generalisation in the context of deep learning.
MuST hosts the CAIR Deep Learning group of the Centre for Artificial Intelligence Research (CAIR), a DSI initiative aimed at developing world-class AI research leadership and teaching capability in South Africa.
Principal Investigator: Dr John Mack
Institution Name: Rhodes University
Active Member Count: 2
Allocation Start: 2022-01-20
Allocation End: 2022-07-21
Used Hours: 21651
Project Name: Rational design of phthalocyanine, porphyrin and BODIPY dyes
Project Shortname: CHEM0796
Discipline Name: Chemistry
The Institute for Nanotechnology Innovation (INI) at Rhodes University under director Prof. Tebello Nyokong carries out research related to the use of molecular dyes in biomedical applications, such as cancer treatment through photodynamic therapy, the use of antimicrobial photodynamic chemotherapy for treating hospital superbugs and as sensors for ions that are harmful to human health.
Other applications of interest include wastewater treatment and the development of optical limiters to protect human vision,such as in aviation safety in protecting pilots from the irresponsible use of laser pointers during runway approaches. It is important to understand trends in the electronic and optical properties that make the molecular dyes suitable for these particular applications. The resources of the Centre for High Performance computing help to facilitate this by making it possible to carry out molecular modelling calculations that can be used to predict how changes to the structures of the molecular dyes will modify their properties. The flexibility that CHPC provides where memory allocations are concerned is often vital.
The INI currently has three staff members, seventeen PhD and ten MSc students, and four postdoctoral fellows, and usually publishes over forty peer-reviewed publications per year. This rose to a high of sixty-five in 2017 with ten involving the use of CHPC resources.
Principal Investigator: Prof Lyudmila Moskaleva
Institution Name: University of the Free State
Active Member Count: 10
Allocation Start: 2022-01-21
Allocation End: 2022-07-22
Used Hours: 937614
Project Name: Understanding surface reactivity of solids using DFT simulations
Project Shortname: CHEM1294
Discipline Name: Chemistry
The group of Prof Moskaleva at the University of the Free State investigates surface reactivity of solids at the atomic level using first-principles quantum-chemical methods, molecular dynamics, statistical theory, microkinetic modelling and thermodynamics. Quantum chemical methods, in particular, those based on density functional theory, are becoming increasingly popular and powerful. They can be used successfully to achieve a mechanistic understanding at the microscopic level, which is needed to optimize functional materials (such as catalysts, electrocatalysts, semiconductors, optoelectronic materials) with respect to their target properties. We would like to mention two of our ongoing subprojects. One of them is a computational study on the chemistry of nanoporous gold, a versatile material possessing interesting mechanical, optical, and catalytic properties. Our study is focused on the remarkable low-temperature catalytic activity of np-Au and its high selectivity toward partial oxidation reactions. CO oxidation has been studied as a model. A dynamic behaviour of the catalytic surface (including the diffusion and restructuring processes) has been studied with the help of ab initio molecular dynamics simulations. We were able to show how O atoms self-organize into chains, a process accompanied by surface coarsening and the formation of an OAuCO intermediate (previously not discussed in the literature) in CO oxidation. In another recently started subproject, we investigate the hydrocarbon combustion chemical reactions using density functional theory (DFT) computational methods. The objectives of this study are to determine valuable information such as geometric, thermodynamic, and kinetic properties of the reactions. Together with our cooperation partners (Prof. Ramakrishnan, TIFR, India) we are performing benchmarking calculations on these system with the high-level W1U, CCSD(T)-DLPNO and various DFT methods with the goal of performance evaluation and choosing an optimal method for future use in generating a training set for Machine Learning. The results will be collected in a dataset MOLDIS, https://moldis-group.github.io/.
Principal Investigator: Prof Michelle Kuttel
Institution Name: University of Cape Town
Active Member Count: 2
Allocation Start: 2022-01-24
Allocation End: 2022-09-21
Used Hours: 188139
Project Name: Molecular modelling of microbial polysaccharides and glycoconjugates
Project Shortname: CHEM1242
Discipline Name: Chemistry
This project involves simulation of carbohydrate molecules using the NAMD simulation software. Atomistic molecular dynamics (MD) simulations provide detailed information on molecular flexibility and dynamics over nanosecond timescales that is currently inaccessible to any experimental method.
Carbohydrate molecules are of particular scientific interest because they play key roles in microbial
infection processes. However, they are challenging to model accurately because they are extremely structurally diverse and very flexible. Although the quality of current additive all-atom force fields for simulating carbohydrate molecules has been demonstrated in many applications, occasional anomalies reported for the behaviour of specific polysaccharides is a cause for concern. A more accurate polarizable "Drude" force field that simulates electronic polarization has been developed to address deficiencies in the current force fields. However, this model has not been tested across a wide variety of simulations.
In this project, we use the NAMD software to run molecular simulations to compare the predictions
of carbohydrate models.
Principal Investigator: Prof Raymond Hewer
Institution Name: University of KwaZulu-Natal
Active Member Count: 8
Allocation Start: 2022-01-24
Allocation End: 2022-08-17
Used Hours: 4499
Project Name: The design and discovery of small-molecule inhibitors of HIV-1 integrase activity
Project Shortname: CHEM0877
Discipline Name: Chemistry
Our research group, led by Prof Raymond Hewer, within the Discipline of Biochemistry at the University of KwaZulu Natal seeks to design and discover novel compounds with therapeutic potential in metabolic disorders that represent a high burden of disease in South Africa. Our primary approach to identifying new compounds involves the use of state-of-art molecular modeling programs which we access through agreement with the CHPC. A number of compounds identified in this manner have proven to be effective in vitro with findings published in various Journal articles
Principal Investigator: Prof David Lokhat
Institution Name: University of KwaZulu-Natal
Active Member Count: 2
Allocation Start: 2022-01-24
Allocation End: 2022-07-25
Used Hours: 11039
Project Name: Carbon dioxide hydrogenation
Project Shortname: CHEM1384
Discipline Name: Chemical Engineering
The Chair in Sustainable Engineering & Process Intensification at UKZN is involved in a number of clean energy and sustainable development projects. Carbon dioxide conversion to useful products is one such project. In this study, the hydrogenation of carbon dioxide over cheap catalysts will be investigated using computational methods. Development of a relatively cheap process for utilizing carbon dioxide will play an important role in our energy transition. The CHPC resources will be used for the computational study, to understand how the various catalytic materials can be used to drive the reaction.
Principal Investigator: Prof Evans Adei
Institution Name: Kwame Nkrumah University of Science and Technology
Active Member Count: 19
Allocation Start: 2022-01-24
Allocation End: 2022-07-25
Used Hours: 2125245
Project Name: Materials For Energy and Fine Chemicals
Project Shortname: CHEM1046
Discipline Name: Chemistry
Mary Mensah in her Ph.D. work at KNUST uses molecular modelling as a tool in investigating the fundamental knowledge required to design the 'best' catalysts for converting lignin (from the cheapest and most abundant source of biomass, agricultural waste for example) to fine/platform chemicals, and as sustainable fossil fuel replacement in the aviation industry in terms of quality and safety.
This research work carried out under the co-supervision of Prof Richard Tia, Prof Evans Adei and Prof Nora de Leeuw is part of the collaborative research programme between the Chemistry Department, KNUST Kumasi, and researchers in the University College London, Cardiff University, University of Botswana, University of Namibia and the University of Johannesburg under the sponsorship of the DFID-Royal Society grant within the framework of Africa Capacity Building Initiative (ACBI). The access to generous computing resources provided by the CHPC was important in Mary's work.
Mary, through lignin models, attempts at developing transition metal catalyst systems that will cleave the pervasive β-O-4 linkages in lignin (de-polymerization process) which constitutes a significant barrier to potential full utilization of lignocellulose biomass. The understanding from this work will also make the high aromatic content of lignin a viable feedstock for the replacement of the existing exploratory technologies that convert aliphatic hydrocarbons (alkanes and alkenes) into aromatic compounds because of the generally high-temperature constraints. The catalysts used in Mary's work also attempt at resolving the challenges of sustainability of the processes which call for the use of environmentally friendly and readily available metals.
CHPC has become an extremely important ally in the training of our students (the future African molecular/material scientists) at KNUST. Our modest contribution to the global understanding of molecular reaction mechanisms in the development of functional materials through our publications would not have been possible without the support of CHPC.
Principal Investigator: Prof Thuto Mosuang
Institution Name: University of Limpopo
Active Member Count: 11
Allocation Start: 2022-01-24
Allocation End: 2022-07-25
Used Hours: 36540
Project Name: Computational studies of various ultra-hard materials
Project Shortname: MATS0875
Discipline Name: Material Science
Currently there are two (2) Doctoral, one (1) Masters, and four (4) Honours students. Computationally the research project investigate various materials such as copper sulphides, copper selenides, gallium nitride, gallium arsenide, graphene oxide, boron nitride, gold and silver nanoparticles. The gold and silver nanoparticles are being researched for possible toxicity/non-toxicity when ingested in human tissues. Specifically, electronic, structural, optical, thermodynamic properties are investigated to enhance semi-conductivity. These properties are then mapped with experimental properties for possible gas sensing and energy materials. DL_POLY software through the CHPC is being utilised in the structural, dynamical, and thermodynamic properties. The exciting code also through the CHPC is being utilised in the electronic and excited properties of copper sulphides and selenides. Now lately, Materials Studio also through CHPC is being used to study interactions between of gold and silver nanoparticles and fibrin protein molecules.
Principal Investigator: Prof Yoshan Moodley
Institution Name: University of Venda
Active Member Count: 4
Allocation Start: 2022-01-25
Allocation End: 2022-08-17
Used Hours: 3759
Project Name: Mammalian Evolutionary Genomics
Project Shortname: CBBI0911
Discipline Name: Bioinformatics
We are based at the University of Venda in the tropical far north of South Africa. My group is involved in several aspects of evolutionary genomics. These include standard bioinformatics, tree building, population genetic and coalescent analyses. reconstructing evolutionary history from whole genome sequences. There is less than a handful of other institutions in Africa that are able to carry out this line of research on wild populations. There are currently several projects being run under this theme by my postgraduate and postdoc researchers. Access to the CHPC is invaluable for our research and the development of my students.
Principal Investigator: Dr Rian Pierneef
Institution Name: University of Pretoria
Active Member Count: 17
Allocation Start: 2022-01-25
Allocation End: 2022-08-17
Used Hours: 89531
Project Name: Bioinformatic and Computational Biology analyses of organisms
Project Shortname: CBBI1124
Discipline Name: Bioinformatics
Programme CBBI1124: Bioinformatic and Computational Biology analyses of organisms is a research programme with foundations in the genomic, transcriptomic and metagenomic analyses of organisms and environments. These research endeavours are based on sequencing data which requires a platform with large computational and memory resources. The CHPC is optimally designed to assist with data analyses and is used on a near daily basis. The projects associated with this programme include but are not limited to food borne pathogen surveillance, wastewater-based epidemiology, antimicrobial resistance surveillance and agriculturally important genomic and transcriptomic profiling. Access to the CHPC enables members of this programme to adequately deliver on their mandates and as such contribute to the health, growth and resilience of South Africa. The projects based under this programme aim not only to contribute to scientific knowledge and outputs but also to the everyday activities of South Africans.
Principal Investigator: Prof Zeno Apostolides
Institution Name: University of Pretoria
Active Member Count: 25
Allocation Start: 2022-01-25
Allocation End: 2022-07-26
Used Hours: 2505
Project Name: Receptomics
Project Shortname: CBBI0926
Discipline Name: Bioinformatics
The Complementary and alternative medicine group in the department of biochemistry, genetics and microbiology at the University of Pretoria work on developing new treatments for diabetes. Diabetes is the second biggest cause of death in South Africa, TB is the first. However, the prevalence of diabetes in increasing. In 2021, 12.8% of the adult population has diabetes, across all ethnic groups has diabetes, but many are unaware of their diabetes status because the symptoms are not visible.
My research aim is to prevent the symptoms of type 2 diabetes with easily available teas, herbs and spices. Many medicinal plants have ethnobotanical claims for diabetes, but few plants have been tested scientifically. We identify such plants in pharmacopoeias from various countries, determine their active ingredients from the literature and test them in computer models of two enzymes that are important in diabetes. We normally start with about 1000 chemical structures in the computer models. Then we select the top 10 that are good inhibitors and are commercially available. We test these 10 with enzyme inhibition assays. Then we calculate how many grams of the tea, herb or spice will be equivalent to the dose of the current drugs for these enzymes. The work on the computer models to screen thousands of chemical structures is done on the super-computers available at the Centre for High Performance Computing (CHPC).
So far, we have found that theoretically, 1 gram of pepper or turmeric or peppermint or rosemary or green tea or purple tea or 3 grams of cinnamon with every meal may inhibit the digestion of starch. These amounts of teas, herbs and spices may be practical solutions for on high starch meals e.g. bread, potatoes, maize or rice to prevent the onset of diabetes. These results suggest that clinical trials are warranted.
These findings may be used to design new drugs for diabetes, in the next 5 - 10 years. These findings may also be useful for plant breeders to develop new cultivars of the teas, herbs and spices that are richer in these anti-diabetic biochemical compounds. Plant breeding and selection is a 10 – 20 year project.
Principal Investigator: Dr Adeola Joseph Oyenubi
Institution Name: University of the Witwatersrand
Active Member Count: 1
Allocation Start: 2022-01-25
Allocation End: 2022-08-11
Used Hours: 26788
Project Name: Optimizing balance in for Impact Evaluation
Project Shortname: ECON1213
Discipline Name: Economics and Finance
This research is conducted by Dr Oyenubi, a Senior Lecturer at the University of the Witwatersrand. My work this period looks at portfolio optimization using heuristic algorithims. The idea is that there are cases where classical optimization methods fails because the assumptions that are required to implement them are not compatible with realistic problems. The research demostrates how such a case can be handled for a portfolio optimization problem. Running code for heuristic optimization ans the associated simmulations often require stronger computing power (in terms of processors) than a desktop machine can handle.
Principal Investigator: Dr Michelle Gordon
Institution Name: University of KwaZulu-Natal
Active Member Count: 4
Allocation Start: 2022-01-25
Allocation End: 2022-08-11
Used Hours: 146812
Project Name: Structural Implications of Mutations in HIV-1
Project Shortname: CBBI0928
Discipline Name: Bioinformatics
Our research group is based at the Nelson R Mandela Medical School at the University of KwaZulu-Natal and focusses on the effects of mutations that cause antiretroviral drug resistance on the structure of HIV-1 proteins. The drugs that patients take as part of their treatment against HIV-1 can stop working because the virus is always changing (mutating). We are looking at how these mutations also change the structure of the virus. To do this, we make a structural model of the affected viral protein using specialised software that uses a lot of computational power. By using the CHPC, I am able to access their high powered server, as well as the software for performing the molecular modelling. With the information gained from this analysis, we can design new drugs that can work against these mutated viruses. The project is progressing well and 5 students have already obtained their degrees using data generated at the CHPC.
Principal Investigator: Prof Yin-Zhe Ma
Institution Name: Stellenbosch University
Active Member Count: 4
Allocation Start: 2022-01-25
Allocation End: 2022-07-26
Used Hours: 2586011
Project Name: Machine learning in 21-cm cosmology
Project Shortname: ASTR1323
Discipline Name: Astrophysics
Chang-Xiang Mao is simulating the epoch of reionization to study the effect of mini-halos (cooling via molecular ways instead of atomic cooling ways in normal halos) in the early universe. Then I use multinest sampling to study the Bayes evidence of the model including mini-halos. Meanwhile, I also calculate the CMB EE_PS and TE_PS.
Principal Investigator: Dr James Sifuna
Institution Name: The Catholic University of Eastern Africa, Nairobi, Kenya
Active Member Count: 7
Allocation Start: 2022-01-26
Allocation End: 2022-08-11
Used Hours: 310261
Project Name: Ab initio study on novel materials for novel functionalities.
Project Shortname: MATS1424
Discipline Name: Material Science
My group, The Theoretical Condensed Matter Group at the Catholic University of Eastern Africa, has two leaders and a couple of students. We have common interests in material discovery. I head the group at the moment and we have had a perfect rapport with every member.
The research group has a very high affinity for new novel materials that are alternatives to fossil energy.
At the moment, we are involved in contributing to the SIESTA methodology. This is evident in the release of Version 5.0 of the code where we helped implement LDA+SOC. In The image, we highlight the success of the implementation by comparing to the well VASP methodology.
CHPC is critical in the sense that it gives us computing facilities we cant achieve within our university. We study large systems that will always need plenty of CPU hours to converge. Thanks to CHPC.
Principal Investigator: Dr Geoff Nitschke
Institution Name: University of Cape Town
Active Member Count: 14
Allocation Start: 2022-01-26
Allocation End: 2022-09-02
Used Hours: 939993
Project Name: Evolving Complexity
Project Shortname: CSCI1142
Discipline Name: Computer Science
Currently consumer products ranging from plastics, to cosmetics to batteries are manufactured from complex compositions of chemical elements and pre-designed chemical substances. An ongoing challenge in the design of new products is to: 1) minimise manufacturing and material costs, 2) minimise the expected environmental impact of the product (during manufacturing and after consumer use), and to 3) only use specific (regulated) materials and chemical compositions. Thus, novel product design can be formulated as a multi-objective optimisation problem. Specifically, where the design of any new product simultaneously satisfies all these constraints, but the product designer is able to manage the weighting (relative importance) of each design objective (constraint). This enables the automated production of a broad array of new products that satisfy the design objectives to varying degrees. A product designer would then ideally select one of several automatically designed products according their own specific constraints for how expensive a product can be, what materials can be used in its manufacture, and what the extent of its expected environmental impact can be. This research investigates multi-objective evolutionary algorithms to automate the design of a vast array of products, given a pre-defined set of materials and chemical substances usable in the design and manufacturing processes, and metrics for expected economic and environmental cost.
Principal Investigator: Dr Nikhil Agrawal
Institution Name: University of KwaZulu-Natal
Active Member Count: 2
Allocation Start: 2022-01-26
Allocation End: 2022-07-27
Used Hours: 98481
Project Name: Molecular Modeling of proteins using different computational tools
Project Shortname: HEAL1189
Discipline Name: Health Sciences
Every function in our body is done by a particular protein. In order to function, protein has to obtain a particular three-dimensional shape. Sometimes proteins obtain a different shape than their normal one, and that leads to disease. It is important to understand misfolding of proteins so that new candidate drug molecules can be designed. In my research, I perform simulations of HIV-related proteins to understand their misfolding.
Principal Investigator: Prof Warren du Plessis
Institution Name: University of Pretoria
Active Member Count: 5
Allocation Start: 2022-01-26
Allocation End: 2022-08-04
Used Hours: 4908
Project Name: Electronic Warfare (EW) Technologies
Project Shortname: MECH0989
Discipline Name: Electrical Engineering
The HPC requirements of this research programme have reduced over the last year. This is likely to change as the processing of large quantities of data is undertaken by postgraduate students.
A paper describing the innovative use of information-theory concepts has been accepted for presentation at a leading international conference. The presented approach seeks to quantify the performance that target-identification algorithms can be expected to achieve even before they are implemented. This means that time is not wasted attempting to implement systems that cannot work or to improve systems that are already performing as well as possible. Simulations performed on the CHPC were essential to this research.
Principal Investigator: Prof Mwadham Kabanda
Institution Name: University of Venda
Active Member Count: 2
Allocation Start: 2022-01-26
Allocation End: 2022-07-27
Used Hours: 104035
Project Name: Reaction mechanism for atmospheric relevant molecules
Project Shortname: CHEM1161
Discipline Name: Chemistry
The research group, under the principal investigator Prof MM Kabanda specializes in the investigation of the reactions mechanisms of molecules with both biological activities and atmospheric influences. Currently Quantum chemical methods are utilised to obtain information on the properties of such molecules
Principal Investigator: Mr Ernest Opoku
Institution Name: Nesvard Institute of Molecular Sciences, Ghana
Active Member Count: 8
Allocation Start: 2022-01-26
Allocation End: 2022-07-27
Used Hours: 35293
Project Name: Molecular Quantum Chemistry
Project Shortname: CHEM1352
Discipline Name: Chemistry
Nesvard Institute of Molecular Sciences is an African-focused private nonprofit research and educational institute in Ghana founded in 2019 and incorporated in 2021 under the companies Act, 2019 (Act 992). The liability of its members is limited by guarantee. Our broader objective is to propagate a new paradigm in molecular science education and research to complement the African development project.
Our vision is to advance molecular sciences in Africa through free and open world class education, training, advocacy, research, and collaboration to prepare the next generation of native African molecular scientists to solve African problems.
We aim to provide fundamental requisite skills that are less common to obtain from traditional educational institutions in Africa to complement further education. We collaborate with laboratory researchers in industry, nonprofit, government, or academic laboratories across Africa and beyond.
And even more than that, our goal has been to teach, mentor and collaborate with younger, but also more experienced native African scientists, on how to set up and perform good scientific research, write good scientific articles, and give them a springboard for further education in rewarding molecular sciences disciplines in more prestigious institutions.
Research works at Nesvard Institute of Molecular Sciences focus on wide area of molecular sciences. We employ techniques of basic and advanced theoretical and computational chemistry and molecular modeling such as (but not limited to) elementary and advanced Hartree-Fock theory, electron correlation methods, density functional theory (DFT), models and concepts of chemistry, linear algebra, symmetry, and group theory, classical and statistical thermodynamics to study molecular properties and associated features.
T