Convergence of sampling in protein simulations

Berk Hess
Phys. Rev. E 65, 031910 – Published 1 March 2002
PDFExport Citation

Abstract

With molecular dynamics protein dynamics can be simulated in atomic detail. Current computers are not fast enough to probe all available conformations, but fluctuations around one conformation can be sampled to a reasonable extent. The motions with the largest fluctuations can be filtered out of a simulation using covariance or principal component analysis. A problem with this analysis is that random diffusion can appear as correlated motion. An analysis is presented of how long a simulation should be to obtain relevant results for global motions. The analysis reveals that the cosine content of the principal components is a good indicator for bad sampling.

  • Received 3 August 2001

DOI:https://doi.org/10.1103/PhysRevE.65.031910

©2002 American Physical Society

Authors & Affiliations

Berk Hess

  • Department of Biophysical Chemistry, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands

References (Subscription Required)

Click to Expand
Issue

Vol. 65, Iss. 3 — March 2002

Reuse & Permissions
Access Options
Author publication services for translation and copyediting assistance advertisement

Authorization Required


×
×

Images

×

Sign up to receive regular email alerts from Physical Review E

Log In

Cancel
×

Search


Article Lookup

Paste a citation or DOI

Enter a citation
×