On Wed, Jul 22, 2015 at 3:25 AM, Albert <mailmd2011.gmail.com> wrote:
> Hello:
>
> I am trying to perform QM/GBSA with mmpbsa.py script. Here is my input
> file:
>
> &general
> endframe=400, interval=20,keep_files=0,
> /
> &gb
> igb=5, saltcon=0.1, qmcharge_com=0, ifqnt=1,
> qmcharge_lig=0,
> qm_residues='1,377,345,249,278,245,347,242,321,337,248,293,285,283,282'
> qm_theory='PM3',
> /
>
> I run the job with command line:
>
> mpirun -np 16 MMPBSA.py.MPI -O -i qmpbsa.in -o QM_MMPBSA.dat -cp
> com.prmtop -rp rec.prmtop -lp lig.prmtop -y pbsa.dcd >log
> However, the job failed with messages:
>
The first thing to do with *any* error you receive running in parallel is
to retry in serial with a shortened example (e.g., try with only one or two
frames). That usually simplifies the error, making it easier to diagnose
what happened.
>
> File "/home/albert/install/amber14/bin/MMPBSA.py.MPI", line 104, in
> <module>
> app.run_mmpbsa()
> File
>
> "/home/albert/install/amber14/lib/python2.7/site-packages/MMPBSA_mods/main.py",
> line 218, in run_mmpbsa
> File "/home/albert/install/amber14/bin/MMPBSA.py.MPI", line 104, in
> <module>
> app.run_mmpbsa()
> File
>
> "/home/albert/install/amber14/lib/python2.7/site-packages/MMPBSA_mods/main.py",
> line 218, in run_mmpbsa
> self.calc_list.run(rank, self.stdout)
> File
>
> "/home/albert/install/amber14/lib/python2.7/site-packages/MMPBSA_mods/calculation.py",
> line 82, in run
> calc.run(rank, stdout=stdout, stderr=stderr)
> File
>
> "/home/albert/install/amber14/lib/python2.7/site-packages/MMPBSA_mods/calculation.py",
> line 157, in run
> self.prmtop))
> CalcError: /home/albert/install/amber14/bin/sander failed with prmtop
> com.prmtop!
> Error occured on rank 8.
>
This says that the 9th CPU (first CPU is rank 0) failed when trying to
compute the QM/MM-GBSA energy with the complex topology file. This is
*all* that the above error message states. The error message generated by
sander will have been placed inside the _MMPSBA_complex_gb.mdout.8 file.
The error message there will hopefully give you more information that will
help solve the problem.
One word of warning is that picking a QM region is difficult if you plan on
allowing bonds to cross the QM-MM boundary. The standard choices allowed
by MMPBSA.py may be inadequate when it comes to picking the QM region, as
it only allows you to pick residues that will be treated with QM. For
protein systems, this will mean that amide bonds -- which are conjugated pi
systems -- will be cut along the QM-MM boundary, which might lead to
difficulties in converging the QM/MM SCF.
If you used exactly the same QM region when you ran dynamics, this is
unlikely the problem. But if you didn't, and you're simply picking a QM/MM
region for the MM-GBSA calculations, you may need to experiment with
different QM regions to get it working (and even then, there is no
guarantee that the results will be any better than for standard MM-PBSA or
MM-GBSA calculations without QM at all). This experimentation will need to
use the -make-mdins and -use-mdins functionality so you can put in your own
custom QM regions. You just have to be careful to make sure the same atoms
are treated with QM in both the bound and unbound states (or the final
binding energies will obviously be meaningless).
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 Wed Jul 22 2015 - 05:00:04 PDT
