Dear Experts,
I am trying to install MPI version of amber14 and ambertool14 to use MMPBSA on gpu cluster here in my university, the details are as follows:
1. The MMPBSA.py.MPI is in $AMBERHOME/bin/MMPBSA.py.MPI
2. While executing the command
$AMBERHOME/bin/MMPBSA.py.MPI -O -i mmpbsa.in -o out.dat -sp solvated.prmtop -cp complex.prmtop -rp receptor.prmtop -lp ligand.prmtop -y last2.5ns.dcd
Loading and checking parameter files for compatibility...
CHAMBER prmtops found. Forcing use of sander
sander found! Using /ghome/atyagiaa/amber14/bin/sander
cpptraj found! Using /ghome/atyagiaa/amber14/bin/cpptraj
Preparing trajectories for simulation...
/d4/atyagiaa/amber14/bin/MMPBSA_mods/make_trajs.py:631: UserWarning: Solvated topology 0 has IFBOX == 0
warn('Solvated topology %s has IFBOX == 0' % ifbox)
Error: Only 'ensemble' mode supported in parallel.
File "/ghome/atyagiaa/amber14/bin/MMPBSA.py.MPI", line 95, in <module>
app.file_setup()
File "/d4/atyagiaa/amber14/bin/MMPBSA_mods/main.py", line 156, in file_setup
self.mpi_size, str(external_progs['cpptraj']), self.pre)
File "/d4/atyagiaa/amber14/bin/MMPBSA_mods/make_trajs.py", line 134, in make_trajectories
traj.Run(pre + 'normal_traj_cpptraj.out')
File "/d4/atyagiaa/amber14/bin/MMPBSA_mods/make_trajs.py", line 684, in Run
raise TrajError('Error running %s' % self.program)
AttributeError: 'Trajectory' object has no attribute 'program'
Exiting. All files have been retained.
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
The error is repeatedly coming, the DO_PARRALLEL used for -np 8, before running the script.
Could you please suggest me what could be the desired solution for this problem.
One more thing I performed MD using NAMD, than converted my files using chamber. One more thing these files easily running on serial version of ambertool14.
Thanks in advance
Abhi
________________________________
From: Abhishek Tyagi <atyagiaa.connect.ust.hk>
Sent: Saturday, March 28, 2015 12:39 AM
To: amber.ambermd.org
Subject: Re: MMPBSA error
Hi,
this is running on gnu cluster on serial installation, I found that for the first problem when i set igb=1, MMGBSA is running, but when it is set igb=5, it shows the error
File "/ghome/atyagiaa/amber14/bin/MMPBSA.py", line 96, in <module>
app.run_mmpbsa()
File "/d4/atyagiaa/amber14/bin/MMPBSA_mods/main.py", line 218, in run_mmpbsa
self.calc_list.run(rank, self.stdout)
File "/d4/atyagiaa/amber14/bin/MMPBSA_mods/calculation.py", line 79, in run
calc.run(rank, stdout=stdout, stderr=stderr)
File "/d4/atyagiaa/amber14/bin/MMPBSA_mods/calculation.py", line 148, in run
self.prmtop))
CalcError: /ghome/atyagiaa/amber14/bin/sander failed with prmtop complex.prmtop!
Exiting. All files have been retained.
for the second problem: I have followed the installation guide as follows:
./configure -mpi gnu
make install
export DO_PARALLEL='mpirun -np 16’
make test.parallel
everything finished fine, but when I had checked AMBERHOME/bin/ their is no MMPBSA.py.MPI file in the folder.
what could be the reason
Thanks in advance
Abhi
~
On 27 Mar, 2015, at 3:00 am, amber-request.ambermd.org<mailto:amber-request.ambermd.org> wrote:
------------------------------
Message: 15
Date: Thu, 26 Mar 2015 08:55:35 -0400
From: Kenneth Huang <kennethneltharion.gmail.com<mailto:kennethneltharion.gmail.com>>
Subject: [AMBER] MMGBSA error
To: AMBER Mailing List <amber.ambermd.org<mailto:amber.ambermd.org>>
Message-ID:
<
CALeh7kBoAEw+CLHBu6USEDxeRu6wYW98Wjy_tyvOnpNpfrLhXA.mail.gmail.com<mailto:
CALeh7kBoAEw+CLHBu6USEDxeRu6wYW98Wjy_tyvOnpNpfrLhXA.mail.gmail.com>>
Content-Type: text/plain; charset=UTF-8
Abhi,
What's the exact error it's giving on the GPU cluster? And have you checked
the _MMPBSA_complex_gb.mdout file to see if there's a description of
what's specifically going wrong?
For your second problem, I'm guessing it's because you haven't compiled
Amber in mpi- you can use the below in the Amber directory.
./configure -mpi gnu
make install
Best,
Kenneth
On Thu, Mar 26, 2015 at 2:06 AM, Abhishek TYAGI <atyagiaa.connect.ust.hk<mailto:atyagiaa.connect.ust.hk>
<javascript:_e(%7B%7D,'cvml',
'atyagiaa.connect.ust.hk<mailto:
atyagiaa.connect.ust.hk>');>> wrote:
Hi,
I have following questions:
1. I am using MMGBSA for receptor-ligand interaction, the calculations
successfully finished on my PC, but when i tried on GPU cluster, I was not
able to run it, for the similar files, it shows error for complex.
2. Second problem is with mpi, i had installed on the gpu cluster
following the installation guidance provided in the Amber manual, and I
have not found MMPBSA.py.MPI on my bin directory.
Could anyone suggest me what could be the error
Thanks? in advance
Abhi
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:AMBER.ambermd.org> <javascript:_e(%7B%7D,'cvml',
'AMBER.ambermd.org<mailto:
AMBER.ambermd.org>');>
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Apr 01 2015 - 07:00:06 PDT
