def run_coverage(bam,
ref_set,
outfile,
tmpdir,
sort=False,
mem=25,
gtmp=4,
docker_image="apollodorus/bioinf:v1"):
tmp = tempfile.NamedTemporaryFile(dir=tmpdir, delete=False)
sort = '#!/bin/bash\nsamtools sort {} > {}'.format(bam, tmp.name)
coverage = 'samtools view -b {} | bedtools coverage -a {} -b - -d > {}'.format(
tmp.name, ref_set, outfile)
exc = tempfile.NamedTemporaryFile(dir=tmpdir, delete=False)
cmd = [coverage]
if sort:
cmd = [sort, coverage]
with open(exc.name, 'w+') as fp:
fp.write('\n'.join(cmd))
os.chmod(exc.name, 0o755)
cmd = bsub(exc.name, mem, gtmp, docker_image, 'genomeCoverage')
po = subprocess.Popen(cmd, shell=True)
return (po)
def run_feature_counts(bam,
saf,
tmpdir,
mem=25,
gtmp=8,
docker_image="genomicpariscentre/subread:1.6.2"):
sample_id = os.path.split(bam)[1].split('_')[0]
fn = sample_id + '_peaks.counts'
outfile = os.path.join(os.path.abspath(tmpdir), fn)
cmd = 'featureCounts -T 5' \
+ ' -a ' + saf \
+ ' -F SAF' \
+ ' -s 0' \
+ ' -o ' + outfile + ' ' \
+ bam
cmd = bsub(cmd, mem, gtmp, docker_image, 'featureCounts')
po = subprocess.Popen(cmd,
shell=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE)
return ((sample_id, outfile), po)
def run_homer(target_set,
bg_set=None,
mem=25,
gtmp=8,
docker_image="apollodorus/homer:mm10"):
# peak_features format:
# 1 - Peak ID, 2- chromosome, 3- start, 4-end, 5-strand
cmd = 'findMotifsGenome.pl '+target_set + ' '\
+ args.build + ' ' \
+ args.output_dir \
+ ' -size '+args.region_size \
+ ' -mis ' + str(args.mismatches) \
+ ' -S ' + str(args.motifs) \
+ ' -p ' + str(args.cpus) \
+ ' -preparsedDir ' + args.preparsed_dir
if bg_set:
cmd = cmd + ' -bg ' + bg_set
cmd = bsub(cmd, mem, gtmp, docker_image, 'annovar', args.debug)
subprocess.check_call(cmd, shell=True)
return (args.output_dir)
def peak_summit_routine(bam, refset, sample_id, tmpdir):
peak_coverage_outfile = os.path.join(tmpdir,
sample_id + '_peakCoverage.bed')
# returns after bedtools coverage complete
get_peak_coverage(bam, refset, peak_coverage_outfile, tmpdir=tmpdir)
exc = os.path.join(os.path.dirname(os.path.abspath(__file__)),
'peakheatmap_summit_bed.py')
peak_summit_outfile = os.path.join(args.output_dir,
sample_id + ".summitCoverage.bed")
cmd = 'python3 {} {} -o {}'.format(exc, peak_coverage_outfile,
peak_summit_outfile)
cmd = bsub(cmd,
mem=30,
gtmp=4,
docker_image='apollodorus/bioinf:pr',
job_name='summitBed')
subprocess.check_call(cmd, shell=True)
# peak_summit_outfile = os.path.join(args.output_dir, sample_id+".summitCoverage.bed")
# get_peak_coverage(bam, tmp.name, peak_summit_outfile, tmpdir)
print('Finished peak summit BED for "{}"'.format(sample_id))
def run_bamCovereage(bam, outfile, mem=25, gtmp=4, docker_image="quay.io/bgruening/galaxy-deeptools"):
tmp = tempfile.NamedTemporaryFile(dir=tmpdir, delete=False)
cmd = 'deeptools bamCoverage -b {} -o {}'.format(bam, outfile)
cmd = bsub(cmd, mem, gtmp, docker_image, 'bamCoverage')
po = subprocess.Popen(cmd, shell=True)
return(po)
def run_bamCoverage(bam,
outfile,
mem=20,
gtmp=4,
docker_image="apollodorus/bioinf:pr"):
cmd = 'bamCoverage --normalizeUsing CPM --binSize 10 -b {} -o {}'.format(
bam, outfile)
cmd = bsub(cmd, mem, gtmp, docker_image, 'genomeCoverage')
po = subprocess.Popen(cmd, shell=True)
return (po)
def run_bam2split(bam, mem=25, gtmp=8, docker_image="apollodorus:dnase2tf:1.0.1"):
sample_id = os.path.split(bam)[1].split('_')[0]
fn = sample_id + 'b.counts'
outfile = os.path.join(os.path.abspath(tmpdir), fn)
cmd = 'featureCounts -T 5' \
+ ' -a ' + saf \
+ ' -F SAF' \
+ ' -s 0' \
+ ' -o ' + outfile + ' ' \
+ bam
cmd = bsub(cmd, mem, gtmp, docker_image, 'featureCounts')
po = subprocess.Popen(cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
return((sample_id, outfile), po)
def run_gene_annotation(avinput,
dist,
build,
db,
outdir,
mem=25,
gtmp=8,
docker_image="apollodorus/annovar:latest"):
prefix = os.path.split(avinput)[1].split('_')[0]
outfile = os.path.join(outdir, prefix + '_annot')
cmd = 'annotate_variation.pl -out '+outfile \
+ ' -build ' + build + ' ' \
+ avinput \
+ ' --geneanno' \
+ ' -neargene '+str(dist) \
+ ' -dbtype refGene ' \
+ db
cmd = bsub(cmd, mem, gtmp, docker_image, 'annovar')
po = subprocess.Popen(cmd, shell=True, stdout=subprocess.PIPE)
return ((po, outfile + '.variant_function'))
parser.add_argument('-o', '--output_dir', default='.', help='')
args = parser.parse_args()
args.peak_bed = os.path.abspath(args.peak_bed)
args.output_dir = os.path.abspath(args.output_dir)
def run_find_motif_bed(find_motif_file):
script = os.path.dirname(os.path.abspath('footprinting_motif_bed.py'))
cmd = 'python3 ' + script \
+ ' ' + args.peak_bed \
+ ' ' + find_motif_file \
+ ' ' + args.fasta
+ ' -o ' + args.output_dir
cmd = bsub(cmd, mem=8, gtmp=4, docker_image='apollodorus/bioinf:pr', 'motif_scan')
print(cmd)
# subprocess.check_call(cmd, shell=True)
def main():
with open(args.peak_motif_list) as fp:
motif_files = fp.read().strip().split('\n')
futures = list()
executor = concurrent.futures.ProcessPoolExecutor(max_workers=60)
for mf in motif_files:
futures.append(executor.submit(run_find_motif_bed, mf))