VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Wed Jul 12 2006 - 11:42:42 CDT

On 7/12/06, John Stone <johns_at_ks.uiuc.edu> wrote:

hi vlad,

> Vlad,
> Even if you hack the PDB plugin to avoid writing out the chain
> identifier, other programs may not be able to read the resulting
> file since many depend on the strictest adherance to the PDB file

this is an often underestimated point. you may get away with
a hacked version of pdb for one tool, but it may fail badly with
the next. the only _consistent_ way of handling it, is to try and
stay as close as possible to the standard. the general rule for
reading writing 'standard' formats is: be permissive on reading
and strict on writing. unfortunately, not all programs adhere to it.

> format specification. The PDB plugin supports reading of various
> illegally formatted PDB files as a result of many MD packages writing
> such files for macromolecules with more than 99,999 atoms and thousands
> of waters, for example. Readding such files into other packages
> is hit and miss, and you can't blame them for not reading them since
> they do indeed violate the PDB format spec. The strategy that
> the PDB plugin is employing once it runs out of the 9999 legal
> residue number limit, is to begin emitting hexadecimal numbers.
> When we exhaust the limit of hexadecimanl numbers, then we begin
> printing asterisks for the residue numbers e.g. "****". If you want
> to alter the behavior of the PDB plugin with regard to these
> out-of-spec files, you can simply change the readpdb.h code and
> modify the write_raw_pdb_record() routine therein, increase the
> plugin version number, recompile, and then load into your VMD and
> you can have it do anything you like. If Leap has an alternative
> way of dealing with PDB files containing > 9999 residues, let me know...

since VMD has a 'predictable' way of handling it, i suggest you write
a small filter program (in awk, perl, python, tcl...), that converts those
illegible 'numbers' to something leap can read and apply it, _after_ you
have generated the files with VMD. since you only want to look at rather
small chunks of the system, a renumbering would not hurt, right?
this way you retain consistent behavior of VMD and get files that work
with leap.

axel.

>
> John Stone
> vmd_at_ks.uiuc.edu
> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  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.