On Sun, Feb 17, 2013 at 12:21 AM, Vaibhav Dixit <vaibhavadixit.gmail.com>wrote:
> Hi,
> We had noticed one behavior that if the ligand is not included in the
> "print_residue" mask, the job for per-residue decomposition analysis never
> seems to end. But with the ligand included it ends normally.
> Is it a bug or we are expected to always include the ligand in the mask?
>
Yes to both. It is a bug, and to avoid it you should _always_ include at
least one receptor residue and at least one ligand residue in the mask.
The reason for this issue is that if no residues are included for either
the receptor or ligand, then MMPBSA.py writes an invalid input file for
that species. (Look at _MMPBSA_gb_decomp_lig.mdin). In the past I only
seen this issue result in a crash as soon as the ligand calculation is
attempted (I have never seen it hang indefinitely).
I will work on a fix for the upcoming AmberTools release, but in the
meantime just make sure that you always pick a ligand residue and a
receptor residue in the print_residues selection.
Thanks for the report,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Feb 17 2013 - 07:30:03 PST
