Abstract
Simulations of molecular systems typically handle interactions within non-bonded pairs. Generating and updating a list of these pairs can be the most time-consuming part of energy calculations for large systems. Thus, efficient non-bonded list processing can speed up the energy calculations significantly. While the asymptotic complexity of current algorithms (namely O(N), where N is the number of particles) is probably the lowest possible, a wide space for optimization is still left. This article offers a heuristic extension to the previously suggested grid based algorithms. We show that, when the average particle movements are slow, simulation time can be reduced considerably. The proposed algorithm has been implemented in the DistanceMatrix class of the molecular modeling package MESHI. MESHI is freely available at 〈www.cs.bgu.ac.il/~meshi〉.
Original language | English |
---|---|
Pages (from-to) | 1041-1048 |
Number of pages | 8 |
Journal | Journal of Computational Biology |
Volume | 13 |
Issue number | 5 |
DOIs | |
State | Published - 1 Jun 2006 |
Keywords
- Energy calculations
- Energy minimization
- Molecular dynamics
- Neighbor list
- Non-bonded list
- Verlet table
ASJC Scopus subject areas
- Modeling and Simulation
- Molecular Biology
- Genetics
- Computational Mathematics
- Computational Theory and Mathematics