VinaLC
VinaLC[1][2] is a parallelized version of Autodock Vina[3]. VinaLC was created by Lawrence Livermore National Lab.
N.B. VinaLC is not being actively developed. It was last updated in September 2018.
Picotte
Installed Versions
VinalC 1.3.0 is installed on Picotte, compiled both with GCC and Intel ICC. Use an appropriate modulefile:
vinalc/gcc/1.3.0
vinalc/intel/2020/1.3.0
Please cite the creators of VinaLC appropriately if you use this software.
Example Job Script
You will need to compare performance between the two ways of running VinaLC below.
MPI only:
#!/bin/bash
#SBATCH -p def
#SBATCH -t 48:00:00
#SBATCH --nodes=2
#SBATCH --ntasks-per-node=48
#SBATCH --cpus-per-task=1
#SBATCH --mem-per-cpu=3G
module load vinalc/intel/2020/1.3.0
${MPI_RUN} vinalc --recList recList.txt --ligList ligList.txt --geoList geoList.txt
Hybrid MPI + threads:
- --ntasks-per-node -- specifies number of MPI ranks (processes)
- --cpus-per-task -- specifies number of threads per rank; we want one CPU core per thread
#!/bin/bash
#SBATCH -p def
#SBATCH -t 48:00:00
#SBATCH --nodes=2
#SBATCH --ntasks-per-node=6
#SBATCH --cpus-per-task=8
#SBATCH --mem-per-cpu=3G
module load vinalc/intel/2020/1.3.0
export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK
export OMP_PROC_BIND=true
${MPI_RUN} vinalc --recList recList.txt --ligList ligList.txt --geoList geoList.txt
Proteus
Installed Version
VinaLC 1.1.2 is installed on Proteus. Use the following modulefile:
vinalc/intel/2015/1.1.2
It is compiled with Intel Composer XE, and Open MPI 1.8.1. For information on running an Open MPI program, see Message Passing Interface#OpenMPI
This will run only on Intel nodes.
Please cite the creators of VinaLC appropriately if you use this software.
Example Job Script
This example job script uses the testcase distributed with the VinaLC source code.
N.B. There are two executables: vinaBMPI, and vinalc. Both seem to
run correctly on the test case, but the documentation does not make
clear the difference between the two. Non-rigorous tests show that the
vinalc program runs slightly faster. It may benefit from hybrid
threading-MPI execution. See: Hybrid MPI-OpenMP Jobs#OpenMPI - Usage
#!/bin/bash
#$ -S /bin/bash
#$ -cwd
#$ -j y
#$ -P FIXME
#$ -M FIXME
#$ -pe fixed16 64
#$ -l ua=sandybridge
#$ -l h_rt=1:00:00
#$ -l m_mem_free=3g
#$ -l h_vmem=4G
. /etc/profile.d/modules.sh
module load shared
module load proteus
module load sge/univa
module load gcc/4.8.1
module load vinalc/intel/2015/1.1.2
# echo "Starting vinaBMPI"
# ${MPI_RUN} vinaBMPI --recList recList.txt --ligList ligList.txt --geoList geoList.txt
# echo "Done vinaBMPI"
echo "Starting vinalc"
${MPI_RUN} vinalc --recList recList.txt --ligList ligList.txt --geoList geoList.txt
echo "Done vinalc"
See Also
References
[2] Xiaohua Zhang, Sergio E. Wong, and Felice C. Lightstone. (2013)
Message Passing Interface and Multithreading Hybrid for Parallel
Molecular Docking of Large Databases on Petascale High Performance
Computing Machines. J. Comput. Chem.