RE: Question about AUTOGENERATE

From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Wed Nov 05 2003 - 12:38:32 CST

Hi,

On Tue, 4 Nov 2003, Zhilei Chen wrote:

> I generate the pdb file through MOE and I manually changed the atom name to
> match the topology file. (the pdb file generated by moe use the atom type as
> the atom name)
> Here I copied a few lines from my pdb file:
>
> -------------------------------
> ATOM 1 C5 BTN 1 0.163 -0.562 -0.563 0.00 C
>
> ATOM 2 H14 BTN 1 0.803 -1.338 -0.892 0.00 H
>
> ATOM 3 C4 BTN 1 0.361 0.680 -1.453 0.00 C

The format of this pdb file is incorrect. Compare one of your lines:

ATOM 1 C5 BTN 1 0.163 -0.562 -0.563 0.00 C

To a file from the PDB databank:

ATOM 1 N ARG 1 12.579 -6.640 11.376 1.00 24.04 6PTI 73

In your file the atom number and name are shifted left two spaces. The
atom number field is ignored, so that appears to work. Also your residue
name is in the correct columns so the segment statement succeeds as well.

However, since the atom name field includes a space to the left of
three-letter names (that is used for four-letter names) psfgen is reading
the name of your first atom as "5" instead of "C5". Therefore the "failed
to set coordinate" warning below is saying that no match is found in your
structure for the atom named "5" in residue from ../biotin2.pdb.

> ---------------------------------------
> Info: generating structure...
> no atom C in residue BTN:1 of segment BIOT
> Warning: add conformation failed in patch NTER
> Info: segment complete.
> reading coordinates from pdb file ../biotin2.pdb for segment biot
> Warning: failed to set coordinate for atom 5 BTN:1 BIOT
> ...............

You should see similar atom name problems if you tried reading this pdb
file into VMD. You might want to ask the MOE developers what format they
are using, and if there is a way to write a standard-conforming file.

-Jim

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