Large-scale Intensive Applications of Molecular Simulations
|Scientific Discipline||Bio / Life Sciences|
|Principal Investigator(s)||Prof. Peter V. Coveney|
|Leading Institution||CCS University College London|
|DEISA Home Site||EPCC|
The Centre for Computational Science, based in the Chemistry Department at the University College of London, has extensive experience in using high performance and Grid computing for large-scale Molecular Dynamics (MD) computations, Recently this has included projects using UK (National Grid Service) and US (TeraGrid) supercomputer facilities, running simulations up to around a million atoms for several nanoseconds of simulated time. The LIAMS project uses that expertise to run parallel MD codes on the DEISA infrastructure to study very large systems. The research has three different strands, taking advantage of the extreme processing power and data storage capabilities provided by DEISA.
In the first application, our goal is to to simulate “life sized” clay platelets which are not constrained by limits on the size of the simulation and thus can reproduce previously unexplored behaviours. We will then address the more complex problems of the interaction of small organic molecules with the clay platelets, and the effect of a single clay sheet upon a polymer matrix.
The other two studies that we propose concern large-scale MD applied to the study of biological macromolecules. In the second application we will investigate the affinity of different inhibitors to HIV-1 protease using exact free energy difference calculations. By harnessing the many supercomputers accessible by DEISA across Europe we will be able to complete a calculation within days instead of the months needed by more traditional methods. In the third application we propose to address the newly emergent area of longmer peptide binding to class I and class II major histocompatibility complexes (MHC).