Re: about charmm topology file

From: Jim Phillips (
Date: Wed Nov 05 2003 - 13:15:48 CST

Hi again,

I suspect the failure to guess coordinates is related to the atom name
misreads due to an odd pdb file format discussed in my previous message.
Psfgen needs to get seed coordinates from somewhere in order to apply the
IC record information to guess hydrogens, missing side chains, etc.

As far as I know, it shouldn't matter if the order of atoms in a pdb file
*used by psfgen* is different than the order in the topology file.

(However, the order of atoms in psf and pdb fed to NAMD or VMD must match
exactly since the atom names in the pdb files are ignored.)


On Wed, 5 Nov 2003, Brian Bennion wrote:
> Second your estradiol residue entry in your topology file should contain
> the atoms in the order they appear in your pdb file for psfgen to have a
> good chance of getting the coordinates correct.

> On Wed, 5 Nov 2003, Zhilei Chen wrote:
> > I am trying to build a charmm topology file for my ligand estradiol. I
> > finished type in the MASS, Charge, Bond, Double, and IMPR. I suppose vmd can
> > auto generate all the angle and bond, based on the pdb file. When I use psfgen
> > to generate pdb and psf, it popped out the message "Warning : failed to set
> > coordinate for atom ..." and "Warning: failed to guess coordinates for 44
> > atoms".

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:05 CST