RE: AMBER: Indentifying cluster center structures in trajectory

From: Ross Walker <ross.rosswalker.co.uk>
Date: Wed, 23 Jul 2008 02:14:27 -0700

Hi Anselm,

I have not tried cluster in ptraj (I use MMTSB to do this and so others may
be able to give a more definitive answer), however, my understanding of
clustering is that the resulting centroids, or cluster centers, are abstract
creations and do not represent a specific structure in your trajectory file.
One could think of them as a type of average structure, or abstract
representative structure. What you are looking for, I believe, is the
structure that has the lowest RMSD to each centroid, this would correspond
to the most "representative" structure in each cluster. This you could
probably do by just running an RMSD fit of your trajectory against each
centroid pdb as a reference structure. The frame with the lowest RMSD in
each case would be the trajectory frame that you refer to below.

Good luck,
Ross

> -----Original Message-----
> From: owner-amber.scripps.edu [mailto:owner-amber.scripps.edu] On Behalf
> Of Anselm Horn
> Sent: Wednesday, July 23, 2008 12:15 AM
> To: amber.scripps.edu
> Subject: AMBER: Indentifying cluster center structures in trajectory
>
> Dear all,
>
> I performed a cluster analysis using ptraj from the AmberTools
> distribution, that worked quite well.
> However, I did not find any hint in the resulting pdb files of the
> cluster centers to what snapshot of the original trajectory they
> correspond.
>
> Is this information printed in any of the other output files?
>
> Thanks in advance,
>
> Anselm
>
> Bioinformatik
> Friedrich-Alexander-Universität
> Erlangen-Nürnberg
> Germany
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
> to majordomo.scripps.edu

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo.scripps.edu
Received on Sun Jul 27 2008 - 06:07:06 PDT
Custom Search