AMBER Archive (2006)

Subject: Re: AMBER: rms in principal component analysis

From: Lydia (aether2006_at_gmail.com)
Date: Wed Jan 18 2006 - 11:43:03 CST


    Thank you very much for your help. I still have two questions about
this. (Sorry that if my questions are too naive.)
(1) Should I use " rms first * " or "rms first out mwtrmsCA.dat :1-52_at_CA"
(in my case, I need to calculate mwcovariance matrix on CA atoms and do pca
only on CA atoms later on)? If both of them are acceptable, which of them
you will choose in my case?
(2) Have the NSCM flag in my MD simulation removed the translation or
rotation already?
sincerely
Jianping
On 1/18/06, David A. Case <case_at_scripps.edu> wrote:
>
> On Tue, Jan 17, 2006, Lydia wrote:
> >
> > Would someone please tell me that if the rms fitting is actually
> needed
> > here in the pca analysis?
>
> It *is* needed: pca analysis (in cartesian coordinates) is not much good
> for
> analyzing overall rotations. If the structure is rotating, the average
> structure becomes pretty meaningless. So, you want to remove all overall
> translation and rotation before proceeding.
>
> ...good luck...dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu