VMD-L Mailing List

From: Axel Kohlmeyer (axel.kohlmeyer_at_theochem.ruhr-uni-bochum.de)
Date: Sat Jan 08 2005 - 13:28:55 CST

>>> "PB" == pbrenne1 <pbrenne1_at_nd.edu> writes:

PB> Hello,

hello paul,

PB> Does anyone know if there is a built-in or plugin module for VMD to do cluster
PB> analysis over a set of dcds? I have used VMD in the past to calculate RMSDs,
PB> but currently I would like to perform conformational cluster analysis as
PB> referenced by Smith, Daura, and Gunsteren in their 2002 Proteins paper. The
PB> analysis is not difficult but a matrix (vector of vectors etc...) would be nice
PB> to store the RMSD values in. From the VMD manual I understand that tcl is not
PB> well suited for mathematical operations such as these. I have also noticed the

it depends on how much mathematics you need. e.g. we have a package for
principal component analysis here, which is currently done completely
in tcl and to make it work reasonably fast for larger systems, we
hope to be able to get away with supplying one(!) lapack call
via an additional binary plugin. also for other analysis tools we have
found, that is is usually sufficient to implement only the key routines
as binary plugins and then do the rest with scripting. writing all
the 'glue-code' is _much_ faster in a scripted language.

alternatively, you can try using the python interface, which should
be much faster for computations.

PB> Cluster visualization plugin contributed by researchers at Cornell but my
PB> initial review found that it required the actual computation be done via a
PB> program written by the Sutcliffe Group in 1996 which does not appear to be open
PB> source.

there is a freely available cluster analysis tool in the gromacs package:
http://www.gromacs.org/documentation/reference_3.2/online/g_cluster.html

the only thing you'd need now is to convert your trajectories to
a format readable by gromacs. incidently, i have just hacked writing
support into the gromacs/trr-plugin for exactly the same reason. ;-)

i'll send in a patch soon, so that the next version of VMD should
have that included. please get back to me, if you want it now.

best regards,
     axel.

PB> Thanks for any responses and my compliments on the VMD framework,
PB> Paul Brenner 

--
=======================================================================
Axel Kohlmeyer       e-mail: axel.kohlmeyer_at_theochem.ruhr-uni-bochum.de
Lehrstuhl fuer Theoretische Chemie          Phone: ++49 (0)234/32-26673
Ruhr-Universitaet Bochum - NC 03/53         Fax:   ++49 (0)234/32-14045
D-44780 Bochum  http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.