On Fri, Feb 13, 2015 at 2:48 AM, Ayesha Fatima <ayeshafatima.69.gmail.com>
wrote:
> Dear All,
> I have run 25ns of simulation using amber 12 and am interested in doing
> hbond analysis of the binding site residues only. I searched round on the
> mailing list and various tutorials and found some files which i modified
> accordingly and used. The input file using residue names ran very well but
> the input with residue numbers as given in Amber tutorial B3 is giving me
> problems. it crashes without any reason at a specific step when it starts
> to process the trajectories.I am attaching both the input and output files.
>
This is something I found in your output file:
Estimated memory usage for this hbond call: 815.12 GB
Unless you're using some kind of Altix machine, there's no way your
machine has that much RAM, so what's happening most likely is ptraj ran out
of memory. (As an aside, reading through the output file can often uncover
enlightening nuggets like this that will suggest an appropriate solution).
You will have to adjust your donor and acceptor masks. In particular:
PTRAJ: donor mask .O
Mask [.O] represents 32742 atoms
This mask selects over 32K atoms, each of which pairs with a donor. You
need to keep the pairs of donor and acceptor masks reasonable for ptraj to
handle.
As an alternative, I highly suggest switching to cpptraj in AmberTools 14
and using the hbond command there. It is quite a bit easier to use (no
need to set up lots of donor and acceptor atom masks). See pages 530 to
532 of the Amber 14 manual for a detailed description of the hbond action,
and feel free to experiment with some of the options to get an
understanding of how it works.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Feb 13 2015 - 02:30:04 PST
