VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Thu May 17 2012 - 07:34:47 CDT

On Thu, May 17, 2012 at 4:19 AM, P.-L. Chau <pc104_at_pasteur.fr> wrote:
> Could I ask about using VMD to split a protein and associated HETATM into
> different PDB files to be used by psfgen to generate files for NAMD
> simulations, please?
>
> I adapted the following script from section 2.1 of the psfgen user's guide
> to split 2bg9.pdb into its consitituent subunits, and it worked well:
>
> package require psfgen
> mol load pdb test.pdb
> set protein [atomselect top protein]
> set chains [lsort -unique [$protein get pfrag]]
> foreach chain $chains {
>    set sel [atomselect top "pfrag $chain"]
>    $sel writepdb nachrchol_${chain}.pdb
> }
> exit
>
> I would like to extend this work. I obtained a PDB file from a colleague
> where there are cholesterol molecules in addition to the 2bg9 protein. I
> would like to split the file into the protein bits and cholesterol
> molecules. I used this same script by executing "vmd -dispdev text -e" on
> the protein+cholesterol system.
>
> However, the procedure crashed with this result:
>
> [...]
> Info)    Atoms: 15608
> Info)    Bonds: 16145
> Info)    Angles: 0  Dihedrals: 0  Impropers: 0  Cross-terms: 0
> Info)    Bondtypes: 0  Angletypes: 0  Dihedraltypes: 0  Impropertypes: 0
> Info)    Residues: 1895
> Info)    Waters: 0
> Info)    Segments: 1
> Info)    Fragments: 22   Protein: 14   Nucleic: 0
> 0
> atomselect0
> -1 0 1 10 11 12 13 2 3 4 5 6 7 8 9
> ERROR) syntax error
> atomselect: cannot parse selection text: pfrag -1
> Info) VMD for MACOSXX86, version 1.9 (March 14, 2011)
> Info) Exiting normally.
>
> where a pfrag has been assigned number -1. I thought I would like to test
> this script on a standard and simpler system, so I downloaded 3dfr.pdb,
> which is dihydrofolate reductase with NADPH and methotrexate bound. I used
> the same script to split the protein and the two HETATM molecules, and I
> obtained this:
>
> Info)    Atoms: 1639
> Info)    Bonds: 1416
> Info)    Angles: 0  Dihedrals: 0  Impropers: 0  Cross-terms: 0
> Info)    Bondtypes: 0  Angletypes: 0  Dihedraltypes: 0  Impropertypes: 0
> Info)    Residues: 428
> Info)    Waters: 264
> Info)    Segments: 1
> Info)    Fragments: 267   Protein: 2   Nucleic: 0
> 0
> atomselect0
> 0 1
> Info) Opened coordinate file nachrchol_0.pdb for writing.
> Info) Finished with coordinate file nachrchol_0.pdb.
> Info) Opened coordinate file nachrchol_1.pdb for writing.
> Info) Finished with coordinate file nachrchol_1.pdb.
> Info) VMD for MACOSXX86, version 1.9 (March 14, 2011)
> Info) Exiting normally.
>
> Only two molecules were written out, when there should be three. I found
> that nachrchol_0.pdb was methotrexate and nachrchol_1.pdb was the protein.
> NADPH was nowhere to be found.
>
> I also ran the protein+cholesterol system on the Tk console interactively,
> and found that the stumbling block is this line:
>
>  set chains [lsort -unique [$protein get pfrag]]
>
> Could I ask if more experienced users would be able to advise me on how to
> solve this problem, please?

VMD does what *you* ask it to do.

your selection assumes that all molecules
that you want to extract classify a protein
according to the corresponding heuristics
in VMD. that obviously cannot work for all
cases. similarly, you assume a valid chain
identifier, however, if your input file may not
follow those conventions.

the basic problem is, that you have limited
information in your input data (pdb file) and
on top of that some of it gets lost during
import, and software has no "common sense".

VMD is deliberately not trying to be too smart,
but rather relies on consistent input data.
so if you want scripts to work well, you have
to make sure that your selections match the
indicators that VMD uses. details on that should
all be listed in the user's guide.

cheers,
     axel.

>
> Thank you very much!
>
> P-L Chau
>
> email: pc104_at_pasteur.fr
> Bioinformatique Structurale
> CNRS URA 2185
> Institut Pasteur
> 75724 Paris
> France
> tel: +33 1 45688546
> fax: +33 1 45688719
> 

-- 
Dr. Axel Kohlmeyer
akohlmey_at_gmail.com  http://goo.gl/1wk0
College of Science and Technology
Temple University, Philadelphia PA, USA.