[gmx-users] pdb2gmx error

[Justin A. Lemkul]

Tsjerk Wassenaar wrote:
> Hi,
> 
>>>        -------------------------------------------------------
>>>        Program pdb2gmx, VERSION 4.0.3
>>>        Source code file: pgutil.c, line: 87
>>>
>>>        Fatal error:
>>>        Atom N not found in residue 1094904186 while adding improper
>>>
>>>        -------------------------------------------------------
>>>
> 
> This could well be something platform/compiler related. I think it
> comes pretty close to the maximum for a signed int.
> 
>>>        If I comment out
>>>         CA    +N     C     O     improper_O_C_X_Y
>>>
>>>
>>>    Right, because if the residue is by itself, there will be no N in
>>>    the next residue (+N).
>>>
>>>
>>> If this is true then pdb2gmx will crash evertime it is run on an isolated
>>> AA, which it doesn't.
>> Not if pdb2gmx recognizes how to treat termini.  If you are using
>> non-standard residues and creating new linkages between polymers, then you
>> may encounter problems.  Perhaps the problem relates the actual residue
>> itself, have you added ASBG to aminoacids.dat?  You haven't showed us your
>> pdb2gmx command line, so we can only assume that something is not being done
>> correctly.
> 
> This is not correct. Okay, you got me doubt myself a bit, so I tested
> and had HOH include a bond from OW to +OW. It's not in either .tdb, so
> according to Justin it would fail by itself. But it doesn't. pdb2gmx
> neglects references to other residues if these residues are not
> present. So something else is wrong. However, for us to be able to
> help, a more complete account of the problem is necessary. The best
> thing to do is to send the .rtp entry and the coordinates, if you're
> not too sensitive about them being archived while your just working on
> them. Also, do provide details regarding version, platform, shoe size
> and anything else that may seem appropriate to diagnose the problem.
> 
> Cheers,
> 
> Tsjerk
> 
> 

Thanks to Tsjerk for pointing out a mistake on my part.  Apologies for being 
inarticulate.  What I was really getting at was that I have seen random errors 
(and not necessarily this one in particular) resulting from not including amino 
acids in aminoacids.dat.  For example, I have used some non-standard amino acids 
in some simulations and seen similar errors when pdb2gmx did not know they 
should be linked like amino acids.  So there are two things I'd suggest 
considering, in addition to the information that Tsjerk suggested:

1. Add ASBG to aminoacids.dat
2. Consider whether or not you need to make an entry into specbond.dat; if the 
polymer is connected through the side chain as you initially said, then you may 
have to define a special bond or atom name to discern this amide bond from the 
one present in the backbone of the amino acid.

-Justin

-- 
========================================

Justin A. Lemkul
Graduate Research Assistant
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================