Hello,
One more comment: I would urge you against changing the input parameters
inside the script itself. That is what the input file is for. We changed
the parallel implementation quite a bit, so that it parallelizes the script
instead of using parallel executables for reasons exactly like this. Not
all MPIs were found to play nice with our old python script (they all play
much nicer now). I've seen this error before, and there was really no way
of fixing it short of installing a different MPI that happened to work. The
new script does still have mpi_cmd and nproc as elements in the INPUT array,
but they are no longer used (one of the reasons I urge you to use the input
file and see the manual for the available options). It was just too much of
a pain to remove them and adjust the indexes.
Also, something I just learned is that 2.6 or later is needed to run
MMPBSA.py (I had been telling people it was 2.5, but I just learned
yesterday that even that is too old).
Good luck!
Jason
On Mon, Jun 14, 2010 at 12:42 PM, Dwight McGee <dwight.mcgee.gmail.com>wrote:
> Hi George,
>
> Judging from your from the email below I would assume that you are using
> one of the older MMPBSA.py scripts. We have made several updates and
> developed a more efficient way to run the MMPBSA.py in parallel
> (MMPBSA.py.MPI). The new program can be downloaded from:
> http://ambermd.org/tutorials/advanced/tutorial3/py_script/compile.htm
> You will need to have a version of python which is 2.5 or later and install
> mpi4py.
> Instructions on how to do this can be found on our website.
> Good Luck!
>
> On Mon, Jun 14, 2010 at 12:10 PM, George Tzotzos <gtzotzos.me.com> wrote:
>
> > Hi everybody,
> >
> > I've attempted to run MMPBSA.py on a parallel amber11 installation.
> >
> > My machine is a quad core iMac running on Intel Core i7
> >
> > I did edit the MMPBSA.py file as follows:
> >
> > 1. Changed INPUT.append('3') to INPUT.append('4')
> >
> > 2. Changed the next line from INPUT.append('none') to
> INPUT.append('mpirun
> > -np 4')
> >
> > The error messages that I get are as follows:
> >
> >
> > g:MMPBSA george1$ MMPBSA.py -O -i mmpbsa_1b.in -o
> FINAL_RESULTS_MMPBSA.dat
> > -sp 2wc6_bom_solv.prmtop -cp 2wc6_bom.prmtop -rp 2wc6.prmtop -lp
> bom.prmtop
> > -y *.mdcrd
> > ptraj found! Using /Users/george1/Programs/amber11_parallel/bin/ptraj
> > sander.MPI found! Using
> > /Users/george1/Programs/amber11_parallel/bin/sander.MPI
> > Warning: igb=2 should be used with mbondi2 pbradii set. Yours are
> modified
> > Bondi radii (mbondi)
> >
> > Preparing trajectories with ptraj...
> > 200 frames were read in and processed by ptraj for use in calculation.
> >
> > Starting sander calls
> >
> > Starting gb calculation...
> >
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Starting pb calculation...
> >
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Error: specified more groups ( 2 ) than the number of
> processors
> > ( 1 ) !
> > application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
> > Error: Sander error! GB mdout file was not created.
> > NOTE: All files have been retained for debugging purposes. Type MMPBSA.py
> > --clean to erase these files.
> >
> > Looking forward to your suggestions how to fix this problems,
> >
> > Best regards
> >
> > George
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
>
> --
> T. Dwight McGee Jr.
> Quantum Theory Project
> University of Florida
> Graduate Student
> dwight.mcgee.gmail.com
>
> "Problems cannot be solved at the same level of awareness that created
> them."
> Albert Einstein
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jun 14 2010 - 11:00:04 PDT
