Search

Home > Research > Computation & Modelling Laboratory
Research

Computation & Modelling Laboratory

The focus of this laboratory is the development of algorithms, modules and packages needed for life science modelling. In the absence of secondary structures for the glycoenzymes identified as targets for cancer and respiratory diseases, we use a combination of homology methods and the SCRU Free Energy Force Induced (FEFI) coarse grain MD to generate the protein structures. These structures are used to develop models for reaction simulations. Following the modelling process, the multidimensional reaction dynamics of complex glycoenzyme catalytic mechanisms are made possible from the SCRU’s development of a Free Energy from Adaptive Reaction Coordinates (FEARCF) Library. Ab Initio quantum dynamics and free energy method make the accurate depiction of  glycoenzyme targets identified from gene expression feature discovery generated in the informatics lab possible.  This accuracy is only possible because the SCRU developed Quantum Supercharger Library (QSL) is capable of rapidly computing wavefunctions for GAMESS-UK, NWChem and other legacy codes. The Computation and Modelling Laboratory’s primary objective is to use these accurate reaction mechanisms to design therapeutics for Cancer and Respiratory diseases. The designed therapeutics from the ab initio quantum level simulations of the enzyme catalysed reactions are tested in the Glycobiomedical laboratory.

 

 

  • Life Science HPC Development Group

    This group’s primary strategy is to enhance the capabilities of molecular simulation codes commonly used by chemists, biophysicists and chemical biologists through the development of functional libraries and modules that link into the legacy codes. The group developed scientific software libraries and collaborate with legacy code developers to create links between our accelerated wavefunction libraries (QSL), Free Energy and Reaction Dynamics (FEARCF) and QM/MM polar bond link atom (SLASH) libraries; firstly with NWChem and then other codes such as GROMOS.  We develop methods for chemical and glycochemical biological applications. Our laboratory designed a new method to model the boundary between quantum and classical compartments for polar (glycosidic bonds) making it possible to accurately compute complex glycans using a mixed quantum classical QM/MM model. We call this the Simple Link Atom Hybrid Saccharide (SLASH) method. The group has developed semi-empirical quantum methods to model chemical glycobiological events that we termed AM1/d-CB1. We develop tools to use a coarse grained potential functions embedded in molecular dynamics calculations to fold and refine protein and glycan conformations. Our specific structural interests are glycosyltransferases, glycosidases and glucokinases. The development of the Free Energy Force Induced (FEFI) coarse grained MD method built on multidimensional free energy profiles These methods developed for the Computational GlycoEnzymology Group make it possible to simulate glycosylation reactions more accurately and design enzyme inhibitors.

    References:

    Werner Crous, Martin J. Field, and Kevin J. Naidoo, Simple Link Atom Saccharide Hybrid (SLASH) Treatment for Glycosidic Bonds at the QM/MM Boundary, J. Chem. Theor. Comput. 2014 10 (4), 1727-1738

    Krishna Govender, Jiali Gao and Kevin J. Naidoo. AM1/d-CB1: A Semiempirical Model for QM/MM Simulations of Chemical Glycobiology Systems. J. Chem. Theor. Comput2014, 10, 4694-4707 Naidoo, K. J., Multidimensional free energy volumes offer unique insights into reaction mechanisms, molecular conformation and association. Phys. Chem. Chem. Phys. 2012, 14, 9026-9036.

    Ananya Gangopadhyay, Simon Winberg and Kevin J. Naidoo, Anisotropic numerical potentials for coarse-grained modelling from high-speed multidimensional lookup table and interpolation algorithms Comp. Chem. 2021 Vol. 42 Issue 10 Pages 666-675

  • Computational GlycoEnzymology Group

    Glycoenzymes (Glycosyltransferases and Glycosidases) are targets for the clinical treatment of several diseases.  In this group we focus on Glycosyltransferases (GTs) that play a central role in cancer i.e., tumour development and Neuraminidases (NANs) central to the infection cycle of respiratory related pathogens.

    We principally used methods developed in the Life Science HPC Group {link} to discover enzyme reaction mechanisms and Transition State (TS) profiles.  This group focuses on investigating the mechanisms of catalytic reactions of GTs and NANs, which process saccharides and alter the glycans on tumour cells and scavenge sialyltransferases from hosts cells. From this a rational design of cancer immunotherapeutics and drugs targeting pneumonia, influenza and coronavirus narrow the testing space for the research group in the Glycobiomedical Laboratory  and biological testing system. We develop from this both Glycomimetics as well as repurposed drug solutions.

    References:

    Barnett, C. B.; Wilkinson, K. A.; Naidoo, K. J., Molecular Details from Computational Reaction Dynamics for the Cellobiohydrolase I Glycosylation Reaction. J. Am. Chem. Soc. 2011, 133, 19474-19482.

    Ian L. Rogers and Kevin J. Naidoo, Multidimensional Reaction Dynamics Reveal How the Enzyme TcTS Suppresses Competing Side Reactions and Their Side Products ,ACS Catalysis 2016 Pages 6384-6392.

  • Computational Chemistry: Molecular Recognition, Conformation and Binding

    Overview: We develop and use computational methods (molecular docking, protein-protein docking, molecular dynamics and free energy simulations), molecular analysis tools (MDAnalysis, MDtraj, Bio3D) and molecular analysis techniques (RMS, time series, correlation functions) to understand molecular recognition, conformation and binding in molecular systems. The combination of these methods is generally applied to biological systems such as molecular machines (enzymes) and to recognition systems at the cellular interface (antibodies). We also contribute to creating tools for repeatable science as part of the open-source Galaxy platform. The aim of the research conducted is to discover and understand molecular recognition events and to assist in prediction and decision making in drug discovery.

     

    Themes: Molecular recognition and interaction, conformation, binding, chemical glycobiology, glycobioinformatics, data analytics, visualisation and repeatable research

    Current projects in 2021:

    • In silico design of MUC1 antibodies to investigate the conformation and binding in MUC1 antigen-antibody systems (Cancer)
    • Screening molecular targets for drug discovery in Malaria

    Continuing projects:

    • Repeatable Large-Scale Data Analysis for Glycobiology and Computational Chemistry

    References:

    Barnett, C. B.;  Senapathi, T.; Naidoo, K. J., Comparative ligand structural analytics illustrated on variably glycosylated MUC1 antigen–antibody binding. Beilstein journal of organic chemistry 2020, 16 (1), 2540-2550.

    Bray, S. A.;  Senapathi, T.;  Barnett, C. B.; Grüning, B. A., Intuitive, reproducible high-throughput molecular dynamics in Galaxy: a tutorial. Journal of Cheminformatics 2020, 12 (1), 54.

  • Computational Chemistry: Ionic Liquids

    Overview: Ionic Liquids (ILs) are broadly defined as molecular salts that are liquid at moderate temperatures, typically consisting of large organic cations and organic or inorganic anions. Low vapour pressures as well as high thermal stability and conductivity provide the motivation for these systems to become next-generation energetic materials and electrolytes.

     

    We develop and use a variety of computational tools to model ionic liquids (ILs) and predict their physical properties.

     

     

     

    Projects include:

    • Using quantum mechanical methods to describe the nature of intermolecular interactions in ILs.
    • The development of next-generation polarisable force fields for ILs.
    • The application of machine learning to predict thermodynamic properties of ILs.

      

  • Laboratory Infrastructure and Resources

    The SCRU hardware platform includes state of the art GPU clusters, data servers and Infiniband clusters.  The compute capability is modest and designed for code development, modelling testing and short simulations prior to using either the national HPC facility (CHPC) or the UCT and regional facility (Ilifu). Our computing environment is set up for computational software developed in-house, from academic developers to commercially licenced. These have been modularized to allow users to manage their environment either in an interactive session or a batch job. Our software modules also enhance users to dynamically change environments user workstations in the computation and modelling lab perform computations on the GPU and CPU cluster computer nodes.

    Computational Cluster:

     Newton: general allocation on nodes - {n1-n4}

     Gibbs: general allocation on nodes - {g1-g8}

     Pauling: general allocation on nodes – {p1-p8}

    GPU Nodes

     Boltz{1-3}

     Kepler

     Kohn

     Tesla