Send to

Choose Destination
See comment in PubMed Commons below
J Phys Chem B. 2006 Nov 16;110(45):22842-52.

Can principal components yield a dimension reduced description of protein dynamics on long time scales?

Author information

Department of Theoretical and Computational Biophysics, Max-Planck-Institute for Biophysical Chemistry, Am Fassberg 11, 37077 Göttingen, Germany.


The suitability of principal component analysis (PCA) to yield slow collective coordinates for use within a dimension reduced description of conformational motions in proteins is evaluated. Two proteins are considered, T4 lysozyme and crambin. We present a quantitative evaluation of the convergence of conformational coordinates obtained with principal component analysis. Detailed analyses of (>200 ns) molecular dynamics trajectories and crystallographic data suggests that simulations of a few nanoseconds should generally provide a stable and statistically reliable definition of the essential and near constraints subspaces. Moreover, a systematic assessment of the density of states of the dynamics of all principal components showed that for an optimal separation of time scales it is crucial to include also side chain atoms in the PCA.

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for American Chemical Society
    Loading ...
    Support Center