starting point for wrapping rail-rna

run_rail_parallel.py

#!/usr/bin/env python

import os
import sys
import subprocess
import time
import logging
import ipyparallel

TEMPLATE="""#!/bin/sh
#SBATCH --ntasks={n}
#SBATCH --mem={mem}
#SBATCH --time={timelimit}
#SBATCH --partition={queue}
#SBATCH --job-name=ipy-engine-{cluster_id}

module load openmpi/1.10.0/gcc-4.9.1
mpiexec -np {n} %s -m ipyparallel.engine --profile-dir="{profile_dir}" --cluster-id="{cluster_id}"
""" % sys.executable

def start_cluster(profile, walltime, partition, mem, n):
    """ Use ipyparallel to start a cluster job. """
    cmd = """ipcluster start \\
        -n {n} \\
        --ip=* \\
        --delay=2 \\
        --profile={profile} \\
        --engines=Slurm \\
        --clean-logs=True \\
        --IPClusterEngines.early_shutdown=100 \\
        --SlurmEngineSetLauncher.batch_template="{template}" \\
        --SlurmEngineSetLauncher.timelimit="{walltime}" \\
        --SlurmEngineSetLauncher.mem="{mem}" \\
        --SlurmEngineSetLauncher.queue="{partition}" \\
        """.format(profile=profile, walltime=walltime, partition=partition, mem=mem, n=n, template=TEMPLATE)

    proc = subprocess.Popen(cmd, shell=True)
    time.sleep(10)
    return proc

def rail(profile):
    cmd = """source activate rail

    rail-rna go parallel \\
        -x ./Drosophila_melanogaster/UCSC/dm3/Sequence/BowtieIndex/genome \\
        ./Drosophila_melanogaster/UCSC/dm3/Sequence/Bowtie2Index/genome \\
        -m https://raw.githubusercontent.com/nellore/rail/master/ex/dm3_example.manifest \\
        -f \\
        --max-task-attempts 10 \\
        --ipython-profile={profile} \\
        #--scratch /home/fearjm/data/railtest/tmp \\
        #--bowtie1 bowtie \\
        #--bowtie1-build bowtie-build \\
        #--bowtie2 bowtie2 \\
        #--bowtie2-build bowtie2-build \\
        #--samtools samtools \\
        #--bedgraphtobigwig bedGraphToBigWig \\
    """.format(profile=profile)
    os.system(cmd)

def stop_cluster(profile):
    os.system(""" ipcluster stop \\
            --profile={profile} \\
            """.format(profile=profile))

if __name__ == '__main__':
    profile='slurm'
    walltime='1:00:00'
    partition = 'niddk'
    mem = '4g'
    n = '60'

    try:
        # Start the cluster
        print('starting cluster')
        cproc = start_cluster(profile, walltime, partition, mem, n)

        # Wait until engines can start
        print('wiating for engines')
        c = ipyparallel.Client(profile=profile)
        while len(c.ids) == 0:
            time.sleep(1)

        print('All OK')
        print(c.ids)

        print('Starting Rail')
        rail(profile)

    except KeyboardInterrupt:
        pass
    finally:
        # Stope the cluster
        print('stoping cluster')
        stop_cluster(profile)