M. L. Dodson wrote:
> Beale, John wrote:
>> I am new to AMBER and I need help with something that I am sure is
>> relatively simple. I am trying to create complexes of small drug
>> molecules with proteins. I have used a docking program to create pdb
>> files of candidate drug-protein complexes. The problem comes in
>> submitting these to AMBER. When I try to run LeAP on the complex pdb
>> file it fails because the portion of the pdb file that represents the
>> ligand doesn’t have the right information. Specifically, the ligand in
>> the pdb file is described by simple atom types without positional
>> designations. For instance, the part of the pdb file representing the
>> ligand gives the atoms as HETATM and may use “C” ten times, instead of
>> “C1, C2, C3, etc.”. It is the same with hydrogens, oxygens, and
>> sulfurs. LeAP complains about this and will not save parameters. Can
>> anyone tell me how I can “fix” my pdb file to contain a model with
>> unique ligand atoms that will run through LeAP? I have tried
>> everything that I can think of, and have never been successful.
>>
>>
>>
>> I have processed my ligand molecules through antechamber and have
>> prepin and frcmod files for each. Will this help me?
>>
>>
>>
>> I am sorry to present such a simplistic problem to this body, but I
>> don’t know how to proceed.
>>
>>
>>
>> John Beale
>>
>>
>>
>
> I don't know of any off the shelf s/w that will do the job, however:
> since you have prepin files, you know the names AMBER assumes the atoms
> to have. So I would proceed from that information to a custom perl
> script to do the renaming. You can verify the correspondence of atom
> number to name by looking at the ligand pdb file in one molecular viewer
> (that allows mouse atom selection) and the prepin (or a pdb file derived
> from it) in another viewer window. From that, create a perl hash
> mapping the atom NUMBER in the ligand file you want to edit to the
> correct atom NAME (from the prepin file). Running this over the ligand
> pdb file should fix it just fine.
>
> You can do the same with awk rather than perl if you know awk better.
>
> You might be able to do it with a LEaP alias table corresponding to the
> perl hash described above, but I have less experience with that. In
> analogy to the hammer and nail saying, to perl hackers, all problems
> look like perl hashes.
>
> If you want to avoid this (possibly reusable for similar problems)
> programming, you can achieve the same thing with an ordinary text editor.
>
> Bud Dodson
Replying to myself, the perl (or awk) approach is made easier by using
pdbcat (since parsing pdb files can itself be a pain).
http://www.ks.uiuc.edu/Development/MDTools/pdbcat/
Bud Dodson
--
M. L. Dodson
Email: mldodson-at-houston-dot-rr-dot-com
Phone: eight_three_two-56_three-386_one
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed May 09 2007 - 06:07:33 PDT
