AMBER Archive (2005)

Subject: RE: AMBER: cluster analysis

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Wed Nov 23 2005 - 10:15:53 CST

Dear Mathew,
 
Concerning the first question, yes you should strip the waters from the
trajectory before doing the clustering otherwise you end up with clusters of
waters biasing the result. Although this may be interesting in some cases,
e.g. you may be able to identify waters or clusters of waters that are
immobile during the simulation. This will take some thinking about an
experimenting with though to find out if it reasonable. If you are just
interested in the solute structure then definitely strip the waters. The
ions you can probably leave as their position within each cluster may be
interesting again, you should experiment with this. Try all 3 situations,
solute+ions+water, solute+ions and then solute and compare the differences.
 
With regards to the options for the kclust command this is part of the mmtsb
toolset and so you need to visit their site to get the manual for the
command.
 
http://mmtsb.scripps.edu/software/mmtsbtoolset.html
 
The kclust documentation if here: http://mmtsb.scripps.edu/doc/kclust.html
 
The radius command defines the extent of each cluster, Essentially how far
away a structure can be from the centroid before it is considered part of
another cluster. It is this option that essentially defines the number of
clusters you will get. The smaller you set the radius value the more cluster
centroids will be generated. I don't believe that one can directly specify
the number of clusters.
 
I hope this helps.
 
All the best
Ross

/\
\/
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel: +1 858 784 8889 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk <http://www.rosswalker.co.uk/> | PGP Key
available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.

 

  _____

From: owner-amber_at_scripps.edu [mailto:owner-amber_at_scripps.edu] On Behalf Of
mathew k varghese
Sent: Tuesday, November 22, 2005 23:03
To: amber_at_scripps.edu
Subject: AMBER: cluster analysis

 Dear AMBER users,

I am trying to do cluster analysis of a 4ns MD simulation of an rna.
I am following the tutorial 8 prepared by Ross Walker. I have some dooubts.

My simulation is in explicit solvent. Should I strip off the WATERS and IONs
before saving pdb files of each frame?

What is the parametr radius used in the kclust command?.

Can we specify the number of clusters?

Thanks in advance
Mathew

/\/\athew

Mathew K Varghese
Research Scholar
School of Pure and Applied Physics
M.G.University
Kottayam, Kerala
India

email mathew_kvarghese_at_yahoo.co.in

===========================================================================

  _____

Enjoy this Diwali with Y! India Click
<http://in.promos.yahoo.com/fabmall/index.html> here

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