[gmx-users] ED analysis

[Berk Hess]

>From: Ruben M Buey <ruben at akilonia.cib.csic.es>
>To: gmx3 at hotmail.com, vdava at davapc1.bioch.dundee.ac.uk
>Subject: Re: [gmx-users] ED analysis
>Date: Mon, 06 Sep 2004 15:18:49 +0200
>
>Berk and Daan,
>Thanks a lot for the reply.
>Just to make sure... (sorry for disturbing). I attach the graph just to 
>check if the "semi-circumference shape" could be due to ramdom diffusion as 
>Berk has just  pointed.

I would say typical random diffusion effect.
Your simulation is probably way to short to see any convergence
for the first few principal modes.
But please read the ref's given in the manual.
You can do two simple checks:
Do PCA on the first and second half of the trajectory and compare
the eigenvectors with g_anaeig to see if there is any convergence.
And check the cosine content of the first few eigenvectors: g_anaeig -cc

Berk.

_________________________________________________________________
Play online games with your friends with MSN Messenger 
http://messenger.msn.nl/