AMBER is a highly parallized suite of biomolecular simulation programs which use MPI and OpenMP on CPUs or CUDA (in combination with MPI) to run on GPUs.
Amber is distributed in two parts: AmberTools and Amber. AmberTools contains a collection of useful software to analyze and modify MD data, whereas Amber is a package of molecular simulation programs.
Running MMPBSA.py.MPI on 20 identical CPUS. Each of the called Sander jobs (used for surface and energy calculation) may take up to 4G of RAM - for bigger Systems this has to be adjusted.
Keep in mind, that MMPBSA calculations are more demanding in terms of memory than MMGBSA jobs.
#!/bin/bash #PBS -l walltime=XX:YY:ZZ #PBS -l select=20:ncpus=1:mpiprocs=1:mem=4G #PBS -r n #PBS -N Job Name #PBS -A Project Name # input files INFILE=MMPBSA.infile EOFILE=Output.dat PRMTOP=Topology.prmtop TRAJ=Trajectory.nc # loas libs module load intelmpi module load Amber/18 # go to work dir cd $PBS_O_WORKDIR # def executable MPI_EXE=`readlink -f ${AMBERHOME}/bin/MMPBSA.py.MPI` # generate log file LOGFILE=$PBS_O_WORKDIR/$PBS_JOBNAME"."$PBS_JOBID".log" echo "$PBS_JOBID ($PBS_JOBNAME) @ `hostname` at `date` in "$PBS_O_WORKDIR" START" > $LOGFILE echo "`date +"%d.%m.%Y-%T"`" >> $LOGFILE echo >> $LOGFILE echo "GLOBAL PARAMETERS">> $LOGFILE echo "---------------------------" >> $LOGFILE echo "Node : "$HOSTNAME >> $LOGFILE echo "Arch : "$ARCH >> $LOGFILE echo "---------------------------" >> $LOGFILE echo "RunDir : "$PBS_O_WORKDIR >> $LOGFILE # MPICPUS=$(cat $PBS_NODEFILE | wc -l) HOSTS=$(cat $PBS_NODEFILE | cut -d"." -f1 | sort -u | tr "\n" ",") echo "# MPI CPUs: ${MPICPUS}" >> $LOGFILE mpirun -n $MPICPUS $MPI_EXE -i $INFILE -eo $EOFILE -cp $PRMTOP -y $TRAJ >>$LOGFILE qstat -f $PBS_JOBID >> $LOGFILE echo "$PBS_JOBID ($PBS_JOBNAME) @ `hostname` at `date` in "$RUNDIR" END" >> $LOGFILE echo "`date +"%d.%m.%Y-%T"`" >> $LOGFILE |