en.plot(plotfile, res = devpars.res, width = devpars.width, height = devpars.height)
if pathview and 'KEGG' in db:
pathviewRDir = path.join(outdir, prefix + '.' + db + '.pathview')
pathviewRfile = path.join(pathviewRDir, 'pathview.R')
shell.mkdir(pathviewRDir)
with open(pathviewRfile, 'w') as f:
f.write("""
{rimport}('__init__.r')
library(pathview)
args = commandArgs(trailingOnly = TRUE)
setwd({pathviewRDir!r})
inopts = {{args.inopts | R}}
inopts$rnames = FALSE
indata = read.table.inopts({infile!r}, inopts)
genes = as.vector(indata[, {genecol}, drop = TRUE])
pvargs = {{args.pathview | R}}
{% raw %}
if (!is.null(pvargs$fccol)) {{
fcdata = as.vector(indata[, pvargs$fccol, drop = TRUE])
names(fcdata) = genes
genes = fcdata
}}
{% endraw %}
pathview(gene.data = genes, pathway.id = args[1], species = 'hsa', gene.idtype="SYMBOL")
""".format(
rimport = rimport, genecol = genecol + 1 if isinstance(genecol, int) else genecol,
infile = infile, pathviewRDir = pathviewRDir)
)
para.run(runPathview, [(pathviewRfile, term.Term.split('_')[-1]) for term in en.results[:top]])
samplevcf = "{{job.outdir}}/{{i.infile | fn}}-%s.vcf" % sample
cmd = '{{args.vcftools}} %s {{i.infile | quote}} > "%s"' % (cmdargs(vtparams), samplevcf)
# vcf2maf.pl --input-vcf ZYYP-ZYYB.vcf --output-maf ZYYP-ZYYB.snpEff.maf --tumor-id ZXLT-ZXLB_TUMOR --normal-id ZXLT-ZXLB_NORMAL --vep-data /path/to/vep/cache/ --filter-vcf /path/to/vep/ExAC_nonTCGA.r0.3.1.sites.vep.vcf.gz --ref-fasta /path/to/hs37d5/phase2_reference_assembly_sequence/hs37d5.fa --vep-path /path/to/miniconda2/bin
params['input-vcf'] = samplevcf
params['output-maf'] = "{{job.outdir}}/{{i.infile | fn}}-%s.maf" % sample
params['vep-data'] = {{args.vepDb | quote}}
params['vep-forks'] = {{args.nthread}}
params['filter-vcf'] = {{args.filtervcf | quote}}
params['ref-fasta'] = {{args.ref | quote}}
params['vep-path'] = path.dirname(vep)
cmd = cmd + '; {{args.vcf2maf}} --tumor-id %s %s' % (sample, cmdargs(params, equal=' '))
cmds.append(cmd)
{% if args.nthread == 1 %}
for cmd in cmds: runcmd(cmd)
{% else %}
# Note the threads may be hanging on here.
p = Parallel({{args.nthread}})
p.run('{}', [(cmd,) for cmd in cmds])
{% endif %}
for i, sample in enumerate(samples):
singlemaf = "{{job.outdir}}/{{i.infile | fn}}-%s.maf" % sample
if i == 0:
runcmd('cat "%s" > {{o.outfile | quote}}' % singlemaf)
else:
runcmd('egrep -v "^#|^Hugo_Symbol" "%s" >> {{o.outfile | quote}}' % singlemaf)
{% endif %}
{% endif %}
invcfs = {{i.infiles | repr}}
outfile = {{o.outfile | quote}}
nthread = {{args.nthread | int}}
joboutdir = {{job.outdir | quote}}
vcftools = {{args.vcftools | quote}}
gatk = {{args.gatk | quote}}
tabix = {{args.tabix | quote}}
ref = {{args.ref | quote}}
params = {{args.params | repr}}
tool = {{args.tool | quote}}
shell.TOOLS.vcftools = vcftools
shell.TOOLS.gatk = gatk
para = Parallel(nthread, raiseExc=True)
invcfs = para.run(vcfIndex, [(vcf, tabix) for vcf in invcfs])
def run_vcftools():
params.d = params.get('d', True)
params.t = params.get('t', True)
params._ = invcfs
params._stdout = outfile
shell.Shell(equal=' ').vcftools(**params).run()
def run_gatk():
params.T = 'CombineVariants'
params.o = outfile
params.R = ref
params.nt = nthread
exts = dict()
for sam in sam_meta:
parts = sam['file_name'].split('.')
ext = '.' + parts[-1]
if ext == '.gz':
ext = '.' + parts[-2] + ext
exts [sam['file_name']] = ext
sample_ids[sam['file_name']] = sam['associated_entities'][0]['entity_submitter_id'][:15]
samfiles = []
for ext in set(exts.values()):
samfiles += glob.glob (os.path.join(os.path.abspath(indir), "*" + ext))
# or direct dir from TCGA download
samfiles += glob.glob (os.path.join(os.path.abspath(indir), "*", "*" + ext))
lock = Lock()
def single(samfile):
bn = os.path.basename (samfile)
if not bn in sample_ids: return
newfile = os.path.join (outdir, sample_ids[bn] + exts[bn])
with lock:
if os.path.exists (newfile):
os.remove(newfile)
if 'link' in method:
os.symlink (samfile, newfile)
elif method == 'copy':
copyfile(samfile, newfile)
p = Parallel(nthread = nthread, backend = 'threading', raiseExc = True)
p.run(single, [(samfile,) for samfile in samfiles])
runcmd(
cmd.format(bedtools=bedtools, params=params, outfile=outfile))
remove(outfile1)
remove(outfile2)
return outfile
infile1 = tabindex(infile1, outdir)
infile2 = tabindex(infile2, outdir)
chroms = [
chr.strip()
for chr in check_output([tabix, '-l', infile1]).splitlines()
]
if nthread > 1:
p = Parallel(nthread, raiseExc=True)
outfiles = p.run(runChrom,
[(infile1, infile2, chrom) for chrom in chroms])
# make sure it's in the right order
outfiles = sorted(outfiles,
key=lambda x: chroms.index(x.split('.')[-2]))
else:
outfiles = []
for chrom in chroms:
outfiles.append(runChrom(infile1, infile2, chrom))
with open(outfile, 'a+') as fout:
for of in outfiles:
with open(of) as f:
fout.write(f.read())
elif tool == 'pyvcf':
import vcf
makedirs(thdir)
asbname = path.basename(affysnps).split('.')[0]
for i, dist in enumerate(dists):
writer = TsvWriter(path.join(thdir, '{bname}.thread{i}.snp'.format(
bname = asbname, i = i
)))
for _ in range(dist):
writer.write(next(reader))
writer.close()
para = Parallel(nthread, raiseExc = True)
para.run(getAlleleCount, [
(tumbam, path.join(
thdir, '{bname}.thread{i}.snp'.format(bname = asbname, i = i)
), path.join(
thdir, '{tumbn}.thread{i}.bamrc'.format(tumbn = path.basename(tumbam), i = i)
)) for i in range(nthread)
])
# merge to tumsnp
writer = TsvWriter(tumsnp)
writer.cnames = ['Chrm', 'pos', 'A', 'C', 'G', 'T', 'Total', 'refCount', 'mutCount']
writer.writeHead(lambda cn: "#" + "\t".join(cn))
for i in range(nthread):
subrc = path.join(
thdir, '{tumbn}.thread{i}.bamrc'.format(tumbn = path.basename(tumbam), i = i)
)
reader = TsvReader(subrc, cnames = False)
for r in reader:
writer.write(r.values())
reader.close()