Re: Cartesian vs dihedral angle space PCA

From: Nicholas M Glykos (
Date: Tue Jul 13 2010 - 04:35:22 CDT

> My question is; is there an emerging consensus in the MD community that
> dihedral PCA is preferable? Should one always present the results from
> both?

Speaking for myself, I think the answer is 'it depends'. For highly
flexible molecules (eg. peptides) where you can't meaningfully separate
rotations/translations from internal motion, dPCA appears to be the best
approach (the papers from the Stock group are highly convincing for that
matter). For structurally stable molecules, Cartesian PCA (especially if
you remove rotations/translations using only the atoms of the structurally
invariant core) is the standard operating procedure. Additionally, the
current implementation of dPCA does not include information about side
chain motion, and it is more difficult to visualise the motion due to
selected eigenvectors. Still speaking for myself, I usually look at both
but only show results from one if they are consistent (stating that
similar results have been obtained from the other). If they are not
consistent it is time to think.

My twopence,

          Dr Nicholas M. Glykos, Department of Molecular Biology
     and Genetics, Democritus University of Thrace, University Campus,
  Dragana, 68100 Alexandroupolis, Greece, Tel/Fax (office) +302551030620,
    Ext.77620, Tel (lab) +302551030615,

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:54:18 CST