Implementation of the replica-exchange molecular dynamics method for rigid bodies
Article first published online: 17 JAN 2006
DOI: 10.1002/qua.20940
Copyright © 2006 Wiley Periodicals, Inc.
Issue
1097-461X/asset/cover.gif?v=1&s=7d5c5cfa0338c787958c2cc55b94246c4766bd2f)
International Journal of Quantum Chemistry
Special Issue: Proceedings of the V. A. Fock Meeting on Quantum and Computational Chemistry
Volume 106, Issue 10, pages 2208–2213, 2006
Additional Information
How to Cite
Moskovsky, A. A., Vanovschi, V. V., Konyukhov, S. S. and Nemukhin, A. V. (2006), Implementation of the replica-exchange molecular dynamics method for rigid bodies. International Journal of Quantum Chemistry, 106: 2208–2213. doi: 10.1002/qua.20940
Publication History
- Issue published online: 25 APR 2006
- Article first published online: 17 JAN 2006
- Manuscript Accepted: 28 NOV 2005
- Manuscript Received: 22 JUL 2005
Funded by
- Russian Foundation for Basic Research. Grant Number: 05-07-90146
- Russian Academy of Sciences (Chemistry and Material Science Section). Grant Number: 10
- Abstract
- Article
- References
- Cited By
Keywords:
- geometry optimization;
- QM/MM methods;
- replica-exchange molecular dynamics (REMD);
- rigid-body molecular dynamics;
- distributed calculations
Abstract
We describe an implementation of the replica-exchange molecular dynamics (REMD) algorithm for rigid-body molecular dynamics targeting its application in the flexible effective fragment quantum mechanical–molecular mechanical (QM/MM) method. The main objective of the stage of the project is to obtain an efficient minimization tool for the MM subsystem. The computer program developed allows one to carry out calculations of molecular dynamics trajectories for atomic particles and for rigid bodies at constant temperature, using the chain Nose–Hoover thermostat. For REMD calculations, the OpenTS dynamic parallelization is used, which allows one to run simulations either on single computational clusters or on meta-clusters. A two-level parallelization scheme has been implemented, assuming that both REMD (upper level) and individual molecular dynamics trajectories (lower level) are parallelized. Structures of small and large water clusters as well as protein conformations are considered as first applications. © 2006 Wiley Periodicals, Inc. Int J Quantum Chem, 2006

1097-461X/asset/QUA_left.gif?v=1&s=c17f602d4ad3c3f31aa9b3acb1b4953adb14cefb)
1097-461X/asset/olbannerright.gif?v=1&s=b363715e04077d5db4c97d27d3e65d17e56e63ef)