With the continuing price reductions and performance increase in personal computer, workstation, and server hardware, computational chemistry researchers are able to develop and use more powerful software to study various theoretical problems and to conduct richer simulations of biochemical processes. The primary efforts we have made include the development and refinement of parallel computing techniques suitable for macromolecular simulation and the construction of the hardware required to efficiently execute it. The latter task includes the evaluation of new hardware technology to ascertain the most cost effective methods of utilizing parallelized codes. Interfacing Q-Chem and CHARMM to perform QM/M reaction path calculations. A hybrid quantum mechanical/molecular mechanical (QM/MM) potential energy function with Hartree-Fock, density functional theory (DFT), and post-HF (RIMP2, MP2, CCSD) capability has been implemented in the CHARMM and Q-Chem software packages. In addition, we have modified CHARMM and Q-Chem to take advantage of the newly introduced replica path and the nudged elastic band methods, which are powerful techniques for studying reaction pathways in a highly parallel (i.e., parallel/parallel) fashion, with each pathway point being distributed to a different node of a large cluster. To test our implementation, a series of systems was studied and comparisons were made to both full QM calculations and previous QM/MM studies and experiments. Furthermore, the recently implemented polarizable Drude water model was used to make comparisons to the popular TIP3P and TIP4P water models for doing QM/MM calculations. We have also computed the energetic profile of the chorismate mutase catalyzed Claisen rearrangement at various QM/MM levels of theory and have compared the results with previous studies. High performance computing hardware and software efforts Construction of a force-decomposition machine for efficient molecular dynamics simulation The seminal work of Plimpton, describes three algorithms for parallelizing molecular dynamics simulations on distributed memory systems: atom decomposition, spatial decomposition, and force decomposition. In collaboration with Urban Borstnik and Dusanka Janezic of the Center for Molecular Modeling, National Institute of Chemistry, Ljubljana, Slovenia, an enhancement to the force decomposition algorithm, Distributed Diagonal Force Decomposition (DDFD), has been developed. LoBoS system resources have been used to build a force decomposition machine to run the DDFD code that has been implemented in CHARMM version c32a2. Results from test runs indicate that this method scales better than the atom decomposition code currently used by CHARMM. Performance comparisons with the spatial decomposition algorithm used by NAMD have also been made. Enabling CHARMM to work on the Open Science Grid One of the most interesting developments in scientific computing over the past decade has been the use of computational grids to perform large-scale calculations. Grid computing has proven useful for biomolecular simulations such as protein folding. Previous non-comprehensive efforts have been made to run CHARMM in a grid environment. Recently, a framework for running CHARMM on the Open Science Grid (OSG) has been developed. The end goal is to allow researchers who have both access to the OSG and a valid CHARMM license to easily run complex simulations on the grid. The workflow management tools developed can also, with some modification, be used with other molecular simulation packages. The utility of the software was demonstrated by a study of the effect of interior hydration on the conformation of the Glu-66 residue of staphylococcal nuclease. Development of techniques and software that will allow CHARMM to run on XMT eXplicit Multi-Therading (XMT) is a general purpose parallel architecture aimed at reducing single task completion time. XMT provides good performance for Parallel Random Access Model (PRAM)-like algorithms, allowing easy programming. We are developing techniques and software so that CHARMM will be used with XMT.
Our research aims to port key concepts of molecular simulation to XMT, including fast Fourier transform, eigenspace operations and short range n-body problem. Our FFT implementation reveals significant performance gaps between XMT and state-of-the-art processors. Future research will be based on implementation of other CHARMM modules, and also realizing molecular simulation sub-programs on XMT. The long-term goal is to completely port CHARMM to XMT.
Showing the most recent 10 out of 15 publications
