From: Axel Kohlmeyer (
Date: Thu Jan 15 2009 - 14:06:16 CST

On Thu, 15 Jan 2009, Javier Goicochea wrote:

JG> Hi all,

hi javier,

JG> I am trying to use correctly topology files. The molecular system I am
JG> building is composed by different substances: solids and liquids, for
JG> example (but not limited to) Quartz and Water or Silicone. So far, I can
JG> create PDB files for the different geometries and substances to be studied.
JG> However, I do not have the topology files for Quartz, nor for the other
JG> solids I want also to study. I am basically interested in using psfgen in
JG> VMD to create the PSF file, but I need the topology file.

there is a topology generation tutorial on the TCBG homepage,
but i don't think this is the proper strategy to address your problem.
the first and most important question that you have to address,
it what kind of interaction potential are you going to be using for
your quartz part of the system and how many of those atoms would
actually be moving. for example, if you need some 3-body potentials,
you will have to do some programming first, IIRC.

then i would recommend to write a code that generates the .psf
file for that crystal part of your system directly alongside
the .pdb file and then you are fine. for the rest of processing
with psfgen you can just read in the existing coordinates and
topology and psfgen should honor it.

JG> At this point, I have been trying to understand the structure / format of
JG> the topology file and I have managed to create a simple (initial) version of
JG> the topology file for Quartz. This topology file can create the PSF file,
JG> with bonds, angles and dihedrals.

the CHARMM type topology (actually "residue template file" would be a
better word) that you want to write implicitly assumes that you want
to build some peptide like polymer. it doesn't work for crystals.
trying to make it work would be much more effort than just writing
your .psf file directly.

but let me reiterate, the most serious problem would be to identify
the force field parameters for quartz, and for all i remember from
what somebody told me a long time ago, you cannot do that well
without 3-body interactions...


Axel Kohlmeyer
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
If you make something idiot-proof, the universe creates a better idiot.