LIN: A new algorithm to simulate the dynamics of biomolecules by combining implicit‐integration and normal mode techniques

Guihua Zhang, Tamar Schlick

Research output: Contribution to journalArticlepeer-review

Abstract

A central goal in molecular dynamics simulations is increasing the integration time‐step to allow the capturing of biomolecular motion on biochemically interesting time frames. We previously made a step in that direction by developing the Langevin/implicit–Euler scheme. Here, we present a modified Langevin/implicit–Euler formulation for molecular dynamics. The new method still maintains the major advantage of the original scheme, namely, stability over a wide range of time‐steps. However, it substantially reduces the damping effect of the high‐frequency modes inherent in the original implicit scheme. The new formulation involves separation of the solution into two components, one of which is solved exactly using normal‐mode techniques, the other of which is solved by implicit numerical integration. In this way, the high‐frequency and fast‐varying components are well resolved in the analytic solution component, while the remaining components of the motion are obtained by a large time‐step integration phase. Full details of the new scheme are presented, accompanied by illustrative examples for a simple pendulum system. An application to liquid butane demonstrates stability of the simulations at time‐steps up to 50 fs, still with activation of the high‐frequency modes. © John Wiley & Sons, Inc.

Original languageEnglish (US)
Pages (from-to)1212-1233
Number of pages22
JournalJournal of Computational Chemistry
Volume14
Issue number10
DOIs
StatePublished - Oct 1993

ASJC Scopus subject areas

  • General Chemistry
  • Computational Mathematics

Fingerprint

Dive into the research topics of 'LIN: A new algorithm to simulate the dynamics of biomolecules by combining implicit‐integration and normal mode techniques'. Together they form a unique fingerprint.

Cite this