def main(opts):
"""Main function"""
org_tfgenes = read_tfgenes(opts.origin)
add_tfgenes = read_tfgenes(opts.addition)
writer = TsvWriter(opts.outfile)
logger.info('Writing the union set to %s ...', opts.outfile)
for gene, tfs in org_tfgenes.items():
for tf in (tfs | add_tfgenes.pop(gene, set())):
writer.write([tf, gene])
for gene, tfs in add_tfgenes.items():
for tf in tfs:
writer.write([tf, gene])
writer.close()
logger.info('Done.')
def getAlleleCount(bamfile, snpfile, outfile):
brcparams = Box()
brcparams.f = ref
brcparams.w = 0
brcparams.l = snpfile
brcparams[''] = bamfile
cmd = '{bamrc} {args} > {outfile!r}'.format(
bamrc = bamrc, args = cmdargs(brcparams, equal = ' '), outfile = outfile + '.tmp')
runcmd(cmd)
# reformated output to desired format
reader = TsvReader(outfile + '.tmp', cnames = False)
snper = TsvReader(snpfile, cnames = False)
#chr1 564773 C 14 =:0:0.00:0.00:0.00:0:0:0.00:0.00:0.00:0:0.00:0.00:0.00 A:0:0.00:0.00:0.00:0:0:0.00:0.00:0.00:0:0.00:0.00:0.00 C:14:... G:0:... T:0:... N:0:...
writer = TsvWriter(outfile)
writer.cnames = ['Chrm', 'pos', 'A', 'C', 'G', 'T', 'Total', 'refCount', 'mutCount']
for r in reader:
while True:
try:
snp = next(snper)
except StopIteration:
break
# use the end position, in case it's 0-based
if snp[0] == r[0] and snp[2] == r[1]:
counts = dict(
A = r[5].split(':', 2)[1],
C = r[6].split(':', 2)[1],
G = r[7].split(':', 2)[1],
T = r[8].split(':', 2)[1]
)
rec = TsvRecord()
rec.Chrm = r[0]
rec.pos = r[1]
rec.Total = r[3]
rec.A = counts['A']
rec.C = counts['C']
rec.G = counts['G']
rec.T = counts['T']
# if reference allele is unknown, assuming all are ref alleles
rec.refCount = counts.get(snp[6].upper(), r[3])
# if mut allele is unknown, assuming no mutations happened
rec.mutCount = counts.get(snp[7].upper(), 0)
writer.write(rec)
# go to next snp
break
else:
# go to next r
continue
writer.close()
random.seed(seed)
reader = TsvReader(infile, cnames=False)
allsnps = set(reader.dump(0))
reader.rewind()
allgenes = set(reader.dump(1))
reader.close()
# assign a probability to each snp
nsnps = len(allsnps)
ngenes = len(allgenes)
snp_probs = dict(zip(allsnps, random.choices(range(ngenes * snppergene),
k=nsnps)))
genebed = TsvWriter(genefile)
snpbed = TsvWriter(snpfile)
geneperchr = math.ceil(float(ngenes) / float(nchr))
for i, gene in enumerate(allgenes):
chrname = 'chr' + str(int(i % nchr) + 1)
start = (int(i / nchr) + 1) * dist
end = start + 1
first_snp_pos = int(start - dist/2.0 - snppergene)
snps = (snp for snp in snp_probs
if i * snppergene <= snp_probs[snp] < (i+1)*snppergene)
genebed.write([chrname, start, end, gene, 0, '+'])
for j, snp in enumerate(snps):
snppos = first_snp_pos + j
snpbed.write([chrname, snppos, snppos, snp, 0, '+'])
# FID, IID, PID, MID, Sex, Pheno
if metafile:
logger.info('Reading metafile ...')
metadata = dict(
TsvReader(metafile, cnames=True, row=lambda r: tuple(
(r.IID, r))).dump())
else:
metadata = None
logger.info('Reading genotype matrix ...')
# snp1 gt1s1 gt1s2 ...
inreader = TsvReader(infile, cnames=True)
samples = inreader.meta[1:]
logger.info('Writing tfam file ...')
tfamWriter = TsvWriter(tfamfile)
tfamWriter.meta = ['FID', 'IID', 'PID', 'MID', 'Sex', 'Pheno']
#tfamWriter.writeHead(callback = lambda meta: '#' + '\t'.join(meta))
if not metadata:
for s in samples:
tfamWriter.write([s, s, '0', '0', 'other', '-9'])
else:
for s in samples:
tfamWriter.write([
metadata[s].FID if s in metadata and 'FID' in metadata[s] else s,
s, (metadata[s].PID or '0')
if s in metadata and 'PID' in metadata[s] else '0',
(metadata[s].MID or '0')
if s in metadata and 'MID' in metadata[s] else '0',
(metadata[s].Sex or 'other')
if s in metadata and 'Sex' in metadata[s] else 'other',
from bioprocs.utils.tsvio2 import TsvWriter, TsvRecord
from gff import Gff
infile = {{i.infile | quote}}
outfile = {{o.outfile | quote}}
attr2name = {{args.attr2name}}
keepinfo = {{args.keepinfo | repr}}
writer = TsvWriter(outfile)
writer.cnames = ['CHR', 'START', 'END', 'NAME', 'SCORE', 'STRAND']
if keepinfo:
writer.cnames.append('ORIGINAL')
def getNameFromAttrs(attrs):
if attr2name:
return attr2name(**attrs)
for key in sorted(attrs.keys()):
if key in writer.cnames:
continue
if 'id' in key.lower():
return attrs[key]
if 'name' in key.lower():
return attrs[key]
return attrs[key]
gff = Gff(infile)
for record in gff:
r = TsvRecord()
r.CHR = record['seqid']
r.START = record['start']
r.END = record['end']
indata1 = TsvReader(infile1, **inopts1) indata2 = TsvReader(infile2, **inopts2) cnames1 = indata1.meta if not rnames1 else indata1.meta[1:] cnames2 = indata2.meta if not rnames2 else indata2.meta[1:] paired = list(set(cnames1) & set(cnames2)) cnames1 = cnames2 = paired if rnames1: cnames1 = [indata1.meta[0]] + cnames1 if rnames2: cnames2 = [indata2.meta[0]] + cnames2 cindex1 = [indata1.meta.index(c) for c in cnames1] cindex2 = [indata2.meta.index(c) for c in cnames2] outdata1 = TsvWriter(outfile1) outdata2 = TsvWriter(outfile2) outdata1.meta = cnames1 outdata2.meta = cnames2 outdata1.writeHead() outdata2.writeHead() for r1 in indata1: outdata1.write(r1[i] for i in cindex1) outdata1.close() for r2 in indata2: outdata2.write(r2[i] for i in cindex2) outdata2.close()
shell.TOOLS.bedtools = bedtools
bedtools = shell.Shell(subcmd = True, dash = '-', equal = ' ').bedtools
params['g'] = gsize
params['i'] = infile
if not 'l' and not 'r' and not 'b' in params:
raise ValueError('You have to define a length to flank (args.params.l, args.params.r or params.b')
if args.extend:
left = params.get('l', params.get('b', 0))
right = params.get('r', params.get('b', 0))
stdns = params.get('s', False)
reader = TsvReader(infile, cnames = False)
writer = TsvWriter(outfile)
for r in reader:
if not stdns or r[5] == '+':
left2, right2 = left, right
else:
left2, right2 = right, left
if params.pct:
length = r[2] - r[1]
r[1] -= round(length * left2)
r[2] += round(length * right2)
else:
r[1] -= left2
r[2] += right2
writer.write(r)
else:
params._stdout = outfile
from bioprocs.utils.tsvio2 import TsvReader, TsvWriter
infile = {{i.infile | quote}}
outfile = {{o.outfile | quote}}
inopts = {{args.inopts | repr}}
infmt = {{args.infmt | quote}}
cutoff = {{args.cutoff | repr}}
degrees = defaultdict(lambda: 0)
if infmt.startswith('pair'):
reader = TsvReader(infile, **inopts)
for r in reader:
if cutoff:
try:
score = float(r[2])
except TypeError:
raise TypeError(
'The 3rd column should be a score for apply the cutoff.')
if score < cutoff:
continue
degrees[r[0]] += 1
degrees[r[1]] += 1
writer = TsvWriter(outfile)
for node in sorted(degrees.keys(), key=lambda x: degrees[x], reverse=True):
if infmt.endswith('complete'):
writer.write([node, int(int(degrees[node]) / 2)])
else:
writer.write([node, degrees[node]])
writer.close()
else:
raise ValueError('Input format other than "pair" not supported yet.')
outfile = {{ o.outfile | quote}}
outdir = {{ o.outdir | quote}}
params = {{ args.params | repr}}
idxfile = {{ args.idxfile | quote}}
kallisto = {{ args.kallisto | quote}}
nthread = {{ args.nthread | repr}}
shell.TOOLS.kallisto = kallisto
params.i = idxfile
params.o = outdir
params.t = nthread
params._ = [fq1, fq2]
kallisto = shell.Shell(subcmd = True).kallisto
kallisto.quant(**params).run()
imfile = path.join(outdir, 'abundance.tsv')
reader = TsvReader(imfile)
writer = TsvWriter(outfile)
writer.cnames = ['target_id', 'est_counts']
writer.writeHead()
for r in reader:
r.target_id = r.target_id.split('::')[0]
try:
r.est_counts = int(round(float(r.est_counts)))
except TypeError:
r.est_counts = 0
writer.write(r)
writer.close()
inopts = {{args.inopts | repr}}
outopts = {{args.outopts | repr}}
refgene = {{args.refgene | quote}}
genecol = {{args.inopts.genecol | repr}}
ocnames = {{args.outopts.cnames | repr}}
del inopts['genecol']
del outopts['cnames']
if region.down is None:
region.down = region.up
# get all genes' TSS and strand
reader = TsvReader(refgene, cnames = False, delimit = '"')
genes = {r[1]:r[0].split("\t")[:7] for r in reader}
reader = TsvReader(infile, **inopts)
writer = TsvWriter(outfile, **outopts)
if ocnames:
writer.cnames = ['CHROM', 'START', 'END', 'NAME', 'SCORE', 'STRAND']
writer.writeHead()
for r in reader:
gene = r[genecol]
if gene not in genes:
msg = 'Gene does not exist: {}'.format(gene)
if notfound == 'error':
raise ValueError(msg)
else:
log2pyppl('Gene does not exist: {msg}', 'warning')
continue
chrom, _, _, start, end, _, strand = genes[gene]
start, end = int(start), int(end)
if strand == '-':
ncol = {{args.ncol}}
nameopt = {{args.name | quote}}
def rowFactory(row):
rparts = row[0].split('_')
if len(rparts) == 4:
(chrom, pos, ref, alt) = rparts
name = chrom + '_' + pos if nameopt == 'neat' else row[0]
elif len(rparts) == 5:
(chrom, pos, name, ref, alt) = rparts
if name == 'NOVEL':
name = chrom + '_' + pos
if nameopt == 'full':
name = row[0]
else:
raise ValueError('Malformat genotype matrix, expect 4 or 5 items in row names.')
if ncol == 3:
return [chrom, pos, int(pos) + 1]
if ncol == 6:
return [chrom, pos, int(pos) + 1, name, 0, '+']
else:
return [chrom, pos, int(pos) + 1, name, 0, '+', ref + ',' + alt, ','.join(rows[1:])]
reader = TsvReader(infile, cnames = True, attach = False, row = rowFactory)
writer = TsvWriter(outfile)
for r in reader:
writer.write(r)
writer.close()
alleles='C,T,',
alleleNs='2634.000000,2374.000000,',
alleleFreqs='0.525958,0.474042,',
bitfields=set(['maf-5-all-pops', 'maf-5-some-pop'])
)
'''
# snps
reader = TsvReader(snpfile, cnames=False)
snplist = list(set(r[snpcol] for r in reader))
reader.close()
from cruzdb import Genome
g = Genome(genome)
outfiletmp = outfile + '.tmp'
writer = TsvWriter(outfiletmp)
for i in range(0, len(snplist), 1000):
chunk = snplist[i:i + 1000]
sql = 'SELECT chrom, chromStart, chromEnd, name, score, strand, refUCSC, alleles, alleleFreqs FROM snp{dbsnpver} WHERE name in ({snps})'.format(
dbsnpver=dbsnpver, snps=', '.join("'{}'".format(s) for s in chunk))
result = g.sql(sql)
for r in result:
allfreqs = dict(zip(r.alleles.split(','),
r.alleleFreqs.split(',')))
reffreq = allfreqs.get(r.refUCSC, '0')
if r.refUCSC in allfreqs:
del allfreqs[r.refUCSC]
if '' in allfreqs:
del allfreqs['']
writer.write([
r.chrom, r.chromStart, r.chromEnd, r.name, r.score, r.strand,
mnames = [m.name for m in MemeReader(tfmotifs)]
motifs = {k:v for k, v in motifns.items() if k in mnames}
logger.info('%s motifs loaded', len(motifs))
if tool == 'meme':
cmdparams = []
params.thresh = pval
params.verbosity = 4
for motif, name in motifs.items():
params.oc = path.join(outdir, name + '.' + re.sub(r'[^\w_]', '', motif))
params.motif = motif
params[""] = [tfmotifs, sfile]
cmdparams.append((meme, cmdargs(params, dash = '--', equal = ' ')))
Parallel(nthread, raiseExc = True).run('{} {}', cmdparams)
writer = TsvWriter(outfile)
writer.cnames = [
"CHR", "START", "END", "NAME", "SCORE", "STRAND", "MOTIF", "SEQ", "STARTONSEQ",
"STOPONSEQ", "RAWSCORE", "PVAL", "QVAL", "MATCHEDSEQ", "UCSCLINK"
]
writer.writeHead(callback = lambda cnames: "#" + "\t".join(cnames))
def rowfactory(r):
r.PVAL = float(r['p-value'])
if r.PVAL >= pval:
return None
r.RAWSCORE = r.score
try:
r.SCORE = int(float(r.score) * 10)
except TypeError:
r.SCORE = 0
try:
reduce
except NameError:
from functools import reduce
return reduce(mul, pvals, 1) ** (1.0/float(len(pvals)))
else:
raise ValueError('Method %s not supported yet.' % method)
def numpval(pval):
try:
return float(pval)
except TypeError:
return 1.0
reader = TsvReader(infile)
writer = TsvWriter(outfile)
prevsnp = None
prevpvals = []
for r in reader:
snp = r.Case.split('.')[0]
if snp != prevsnp:
if prevsnp:
writer.write([
prevsnp,
aggregate(prevpvals, method)
])
prevsnp = snp
prevpvals = [numpval(r.Pval)]
else:
prevpvals.append(numpval(r.Pval))
writer.write([
open(outdata, 'w').close()
open(outgroup, 'w').close()
open(outcase, 'w').close()
exit(0)
# save the data file
# expfile
"""
S1 S2 .. Sn
G1 ...
G2 ...
"""
expreader = TsvReader(expfile)
expdata = [r for r in expreader if r[0] in genes or r[0] in tfs]
expreader.close()
datawriter = TsvWriter(outdata)
for i, cname in enumerate(expreader.cnames):
if i == 0:
# genes + tfs
datawriter.cnames = [r[0] for r in expdata]
datawriter.writeHead()
else:
datawriter.write([cname] + [r[i] for r in expdata])
datawriter.close()
del expdata
genes = [g for g in genes if g in datawriter.cnames]
tfs = [g for g in tfs if g in datawriter.cnames]
genetfs = {g: [tf for tf in gtfs if tf in tfs] for g, gtfs in genetfs.items() if g in genes}
# save the group file
writer.writeMeta()
for _ in range(joblist[i]):
try:
writer.write(reader.next())
except StopIteration:
break
writer.close()
thparams = params.copy()
thparams[""] = [qfile, mfile2]
thparams.thresh = qval
thparams.oc = ocdir
cmdps.append((tomtom, cmdargs(thparams, dash='-', equal=' ')))
reader.close()
Parallel(nthread, raiseExc=True).run('{} {}', cmdps)
writer = TsvWriter(outfile)
reader = TsvReader(path.join(ocdirs[0], 'tomtom.txt'),
comment='##',
cnames=lambda header: header[1:].strip().split("\t"))
writer.cnames = reader.cnames
writer.writeHead(lambda cnames: "#" + "\t".join(cnames))
reader.close()
for ocdir in ocdirs:
reader = TsvReader(
path.join(ocdir, 'tomtom.txt'),
comment='##',
cnames=lambda header: header[1:].strip().split("\t"))
for r in reader:
writer.write(r)
reader.close()
writer.close()
getAlleleCount(normbam, affysnps, normsnp)
else:
# try to split the affysnps into N files and distribute the jobs to nthreads
# get number of lines of affysnps file
total = wc_l(affysnps)
dists = distribute(total, nthread)
reader = TsvReader(affysnps, cnames = False)
# dir to save the split file and result file
thdir = path.join(outdir, 'bamrc.nthreads')
if not path.exists(thdir):
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)
