[gmx-users] DCCM

[Mark Abraham]

tuhin at iitk.ac.in wrote:
> Dear Gromacs Users,
> 
> My apologies for putting forward a naive question.
> 
> Actually, I'm a little confused, is covariance matrix the same as
> correlation matrix ??

Names for such procedures vary quite a bit with the kind of field you 
work in - I've known scientists, engineers and statisticians to use 
different names for the same arithmetic process. In the absence of some 
unequivocal mathematical expression for what you're trying to do, it's 
hard to guess what you're doing, unfortunately.

I can see below you've been doing some suitable reading, so perhaps you 
should try to come up with such a mathematical description of what 
you're trying to get your analysis to do. That may give you enough 
insight to see how to do it with a GROMACS tool or some other.

Mark

> I had been thinking of carryingout DCCM analysis for
> my trajectories (20ns each), one for a WT protein and the other a mutant
> version. I thought of carrying out the analysis to figureout what are the
> residue pairs (if any??) that show correlated/anti-correlated motion
> during the course of the MD. Even if I talk about residue pairs I would be
> taking the Ca-atoms for each residues. I have checked few articles also:
> 
> 1. Proteins, 11:205-217 (1991)
> 2. BMC struct. Biol., 7:73 (2007)
> 3. Proteins, 65:843-855 (2006)
> 
> (Though, in article:2 and 3 they have used inhouse codes for DCCM
> construction)
> 
> As had been discussed long back in the mailing list (dt: 18 Sept 2007,
> raised by Dhananjoy Joshi) people suggested about using:
> 
> g_covar -xpma (-xpm) OR g_covar -ascii (using the diagonal coponents etc...)
> 
> but for my analysis I find it really difficult to carryout the analysis or
> to find any significance coming out of my analysis using g_covar (might be
> I'm doing it wrong!!). Thus, I felt I would post my mail in the gmx-list
> to invite suggestions from the veterans.
> 
> I would really appreciate if someone could guide me carryout the DCCM
> analysis properly OR share his expertise.
> 
> Warm regards,
> Tuhin
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>