Re: Analysis of NAMD output with CHARMM

From: \ (
Date: Mon Feb 14 2005 - 12:15:58 CST

   Hello. I have done this before, and I can tell there is definitely a problem
with your psf file. Unlike NAMD or psfgen, CHARMM defines "groups" of atoms
whose total charge equals zero. Thus, you would have at least as many groups as
residues. Nevertheless, according to your output:

>Number of segments = 11 Number of residues = 12080
>Number of atoms = 39548 Number of groups = 1
>Number of bonds = 39590 Number of angles = 19283
>Number of dihedrals = 10881 Number of impropers = 678
>Number of HB acceptors = 0 Number of HB donors = 0
>Number of NB exclusions = 0 Total charge = 0.00000

there is only one group in which 12080 residues are contained, thus the error:

> ***** LEVEL -3 WARNING FROM <MAKGRP> *****
> ***** maximum number of atoms per group exceeded
> ******************************************

   This is actually a very common error: CHARMM psf files have a section called
something like "NGRP NST2" which defines the system's groups, while psfgen
always leaves this section blank. The best (possbly sanest) way of overcoming
this is generating the psf file directly in CHARMM. If you have never done this
before you can check "gen-prot.inp" from the Script Archive on CHARMM's forum
( Remember: when generating segments in CHARMM you must read
each from a different pdb file.

   I would also suggest generating the coordinate file (crd file) also in
CHARMM, but beware: this file must preserve the exact order of the atoms in
NAMD's pdb file (which is the same order in your trajectory file). This is very
important because the dcd file has no atom names, but only atom numbers.

   I hope this helps.

J. Valencia.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:39:10 CST