Re: Re: seg fault and namdplot problem

From: Jim Phillips (
Date: Mon Sep 22 2003 - 18:20:20 CDT

Hi Cheri,

First, psfgen guesses H positions based on the IC records in the topology
file or, failing that (since there are no IC records for H atoms in the
top_all22_prot.inp HEME residue), with some very crude rules assuming a
tetrahedral configuration. The goal is to get close enough for the
minimizer to work. If you use VMD with both the psf and pdb files you'll
see the real bonds, but don't make any single-bond assumptions based on
the initial geometry.

Regarding namdplot, bash has a very different syntax than csh (bash is
compatible with sh, but only steals features from ksh and csh). If you
don't have csh you may have tcsh. If it's not in /bin try /usr/bin/ or
/usr/local/bin (or the unix trick "/bin/env tcsh").


On Mon, 22 Sep 2003, Cheri M Turman wrote:

> Hi Jim,
> You were right about the files I had sent you. I had tried so many
> things that I couldn't really tell what went with what. However, I saw
> the H coordinates as you said in the pdb file and all were 0. The seg
> fault was due to these coordinates being 0. I added guesscoor after the
> apoprotein segment and heme segment as you suggested and the H then got
> coordinates assigned. The problem with this is that the heme is
> distorted as seen in vmd or SPDBViewer. My original psfgen had the
> guesscoord at this spot which is why they worked before in namd. But
> when I tried to prevent the heme distortion, I didn't allow psfgen to
> guees coordinates of the heme only the apoprotein. This is when the 0.0
> are generated for the H of the heme. In the distorted heme, it appears
> as if psfgen is putting hydrogens in a frame where the carbons in the
> heme ring (not part of pyrrole rings, ie. CHC) are sharing the hydrogen
> with the adjacent carbon in the pyrrole ring (ie. C4B). There should be
> a double bond between these two carbons anyway leaving CHC with no room
> for an H. I cannot figure out why this is occuring. Also, all other H
> (ie. on methyl groups) also form bonds to each other. These are both
> puzzling me. I have just started to manually modify the H to prevent
> these bonds in the viewers. Does anyone have comments on these problems?
> Also, Jim, I am trying to run your namdplot script but I keep getting
> syntax errors. I had to change /bin/csh to /bin/bash so I was wondering
> if my interpreter is not understanding the syntax. I run Debian and the
> bash is suppose to read also csh. The error looked like this:
> ./namdplot: line 37: syntax error near unexpected token `('
> ./namdplot: line 37: ` foreach t ( $titles )'
> Removing the "()" got me further down the program but I had to keep
> removing these. Finally, at the end it doesn't like how the file ends
> so I get this error:
> ./namdplot: line 140: syntax error: unexpected end of file
> Please let me know your thoughts...should I install an original csh?
> Thanks,
> Cheri
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Cheri M. Turman
> Graduate Student
> University of Texas-Houston Medical School
> 6431 Fannin
> Houston, TX 77030
> e-mail:
> Ph.: 713-500-6126
> Fax: 713-500-0652

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:02 CST