Computational Modelling Group

Seminar  22nd March 2011 4 p.m.  27/2003

Computational biochemistry: understanding biomolecular mechanisms by modelling

Professor Adrian Mulholland
School of Chemistry, University of Bristol

Web page
http://www.chm.bris.ac.uk/pt/ajm/Site/Home.html
Categories
AMBER, Biomedical, Biomolecular Organisation, Biomolecular simulations, C, CASTEP, Complex Systems, Computer Science, Energy, FFT, Finite elements, Fortran, Gaussian, HECToR, HPC, HPCx, Iridis, Jaguar, Linux, Molecular Dynamics, Molecular Mechanics, Monte Carlo, MPI, MS Office Access, Multi-physics, Multi-scale, Multipole methods, Onetep, Optimisation, ProtoMS, Quantum Chemistry, Scientific Computing, Software Engineering, Structural biology, Systems biology, Xmgrace
Submitter
Chris-Kriton Skylaris

Professor Adrian Mulholland

Enzymes are outstandingly efficient natural catalysts. Knowledge of the principles underlying their catalytic properties promises technological spin-offs such as routes to new drugs (e.g. in the design of enzyme inhibitors, such as analogues of transition states or reaction intermediates); analysis of the effects of genetic variation and mutation (e.g. in predicting the effects of single nucleotide polymorphisms on drug metabolism); design of new catalysts (e.g. engineered enzymes or designed biomimetic catalysts) for industrial applications. Molecular simulations offer the potential of uniquely detailed understanding of enzyme catalytic mechanisms. Combined quantum mechanics/molecular mechanics (QM/MM) methods are a good approach to modelling enzyme catalytic mechanisms. Recent applications include studies of drug metabolizing enzymes such as cytochrome P450 and beta-lactamases (involved in antibiotic resistance), e.g. investigating the effects of mutations, identifying catalytic interactions, and modelling complexes with drugs. QM/MM modelling of fatty acid amide hydrolase (FAAH) reveals complex conformational effects in catalysis, and has also identified the productive binding mode of promising covalent inhibitors, showing the potential of QM/MM methods to contribute to drug design. The effects of quantum tunnelling can be analysed, e.g. in studies of aromatic amine dehydrogenase. Classical molecular dynamics simulations allow large scale protein conformational changes to be investigated, and can identify functionally relevant motions. New techniques also allow high level QM/MM calculations of relative free energies. Calculations using high-level QM/MM methods can now predict barriers for enzyme-catalysed reactions with unprecedented accuracy. Quantitative predictions from first-principles calculations were only previously possible for very small molecules. Calculated activation energies for enzymic reactions, such as that catalysed by chorismate mutase, are in excellent agreement with experiment. With these methods, quantitative, reliable predictions can be made about the mechanisms of enzyme-catalysed reactions. This development opens new horizons for computational biochemistry and enzymology.