AMBER Archive (2005)

Subject: Re: AMBER: ambpdb TER problems

From: samuel.arey_at_epfl.ch
Date: Mon Dec 26 2005 - 13:26:27 CST


Thanks very much for your input.
Ok I understand my oversight about the pdb file
format.

>Are you sure that the names and numbers of the input pdb file around this
>region are fully correct?

Well, the input pdb file has incomplete side chains,
and then of course xleap automatically "built" these,
in the sequence of commands shown farther down.

I am an AMBER novice, but I carefully double-checked
that the sequence and names of atom types in the
xleap-modified .lib file followed those found in the
standard libraries. They looked fine -- at least in the
CSER, DA5, and adjacent residues I am discussing here.
When I used "savepdb" to generate a pdb file of the
xleap-modified structure, these residues looked ok
and were separated by a TER card in the right place.

From a "cleaned up" pdb file I created a solvated
structure using the standard commands:

 omc = loadpdb protein.pdb
 check omc
 addions omc K+ 0
 solvatebox omc TIP3PBOX 8.0
 saveoff omc omc.lib
 saveamberparm omc omc.top omc.crd

The resulting prmtop file can be temporarily downloaded
from my website:

http://lcbcpc21.epfl.ch/Group_members/sam/omc.top

Sam

David A. Case wrote:

>On Fri, Dec 23, 2005, Sam wrote:
>
>>and found several problems in the newly created pdb file:
>>(1) The terminal protein residue (CSER) is incorrectly
>>identified as nonterminal ("SER").
>
>
>This is not incorrect: the pdb format does not use special resdiue names
>for beginning or terminal residues. Hence, "SER" is correct.
>
>>(2) No TER card is inserted
>>between this residue (SER) and the starting terminal residue
>>(DA5) of the first DNA strand. (3) A TER card is incorrectly
>>inserted partway into the DA5 residue.
>
>
>This is weird, and I have not seen this behavior before. I suspect you will
>have to post your prmtop file, or the procedure you used to create it.
>Are you sure that the names and numbers of the input pdb file around this
>region are fully correct? This kind of looks like a bug in ambpdb, since
>it should not put a TER card in the middle of a residue. But if the prmtop
>file is bad somehow, it could be fooled.
>
>....dac
>
>-----------------------------------------------------------------------
>The AMBER Mail Reflector
>To post, send mail to amber_at_scripps.edu
>To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu