def detectBp(range1, range2, blockSize, assembly='hg19'):
# process coordinate
range1 = mygenome.locus(range1)
range2 = mygenome.locus(range2)
# fetch sequence
seq1 = range1.nibFrag('/data1/Sequence/ucsc_%s' % assembly, blockSize,0)
seq2 = range2.nibFrag('/data1/Sequence/ucsc_%s' % assembly, 0,blockSize)
# generate block combination
outFile = open('/EQL2/RNASeq_LymphNK/bpTemplate.fa','w')
for i in range(len(seq1)-blockSize+1):
for j in range(len(seq2)-blockSize+1):
outFile.write('>seq1:%s-%s|seq2:%s-%s\n%s%s\n' % (i,i+blockSize,j,j+blockSize, seq1[i:i+blockSize], seq2[j:j+blockSize]))
outFile.close()
# run bowtie
print 'building bowtie index'
os.system('bowtie-build -q -f /EQL2/RNASeq_LymphNK/bpTemplate.fa /data1/Sequence/bowtie/bpTemplate')
def process_bp(inFileName,outFileName,regionL):
result = mygsnap.gsnapFile(inFileName,True)
outFile = open(outFileName, 'w')
outFile.write('browser full knownGene\n')
outFile.write('track name="%s" visibility=2\n' % inFileName)
for rL in result:
if not (rL[0].nLoci==1 and rL[1].nLoci==1 and rL[0].pairRel=='concordant'):
raise Exception
locL = [mygenome.locus(rL[0].matchL()[0].segL[0][2]), mygenome.locus(rL[1].matchL()[0].segL[0][2])]
flag = False
for loc in locL:
for region in regionL:
if loc.overlap(region) > 0:
flag = True
if flag:
print '^%s.*%s$\n' % (rL[0].seq(),mybasic.rc(rL[1].seq())),
for loc in locL:
outFile.write('%s\t%s\t%s\n' % (loc.chrom,loc.chrSta,loc.chrEnd))
def process_bed(inFileName, outFileName, coordH):
result = mygsnap.gsnapFile(inFileName, True)
outFile = open(outFileName, 'w')
count_all = 0
count_strand = 0
outFile.write('browser full knownGene\n')
outFile.write('track name="targeted" visibility=2\n')
for rL in result:
if not (rL[0].nLoci == 1 and rL[1].nLoci == 1
and rL[0].pairRel == 'unpaired'):
raise Exception
locL = [
mygenome.locus(rL[0].matchL()[0].segL[0][2]),
mygenome.locus(rL[1].matchL()[0].segL[0][2])
]
for loc in locL:
loc.chrSta += coordH[loc.chrom][1] - 1
loc.chrEnd += coordH[loc.chrom][1] - 1
loc.chrom = coordH[loc.chrom][0]
for loc in locL:
outFile.write('%s\t%s\t%s\n' % (loc.chrom, loc.chrSta, loc.chrEnd))
def process_bed(inFileName,outFileName,coordH):
result = mygsnap.gsnapFile(inFileName,True)
outFile = open(outFileName, 'w')
count_all = 0
count_strand = 0
outFile.write('browser full knownGene\n')
outFile.write('track name="targeted" visibility=2\n')
for rL in result:
if not (rL[0].nLoci==1 and rL[1].nLoci==1 and rL[0].pairRel=='unpaired'):
raise Exception
locL = [mygenome.locus(rL[0].matchL()[0].segL[0][2]), mygenome.locus(rL[1].matchL()[0].segL[0][2])]
for loc in locL:
loc.chrSta += coordH[loc.chrom][1] -1
loc.chrEnd += coordH[loc.chrom][1] -1
loc.chrom = coordH[loc.chrom][0]
for loc in locL:
outFile.write('%s\t%s\t%s\n' % (loc.chrom,loc.chrSta,loc.chrEnd))
def process_bp(inFileName,outFileName,coordH,regionL):
result = mygsnap.gsnapFile(inFileName,True)
outFile = open(outFileName, 'w')
outFile.write('browser full knownGene\n')
outFile.write('track name="%s" visibility=2\n' % inFileName)
for rL in result:
if not (rL[0].nLoci==1 and rL[1].nLoci==1 and rL[0].pairRel=='unpaired'):
raise Exception
locL = [mygenome.locus(rL[0].matchL()[0].segL[0][2]), mygenome.locus(rL[1].matchL()[0].segL[0][2])]
for loc in locL:
loc.chrSta += coordH[loc.chrom][1] -1
loc.chrEnd += coordH[loc.chrom][1] -1
loc.chrom = coordH[loc.chrom][0]
flag = False
for loc in locL:
for region in regionL:
if loc.overlap(region) > 0:
flag = True
if flag:
print '^%s.*%s$\n' % (rL[0].seq(),mybasic.rc(rL[1].seq())),
for loc in locL:
outFile.write('%s\t%s\t%s\n' % (loc.chrom,loc.chrSta,loc.chrEnd))
def detectBp(range1, range2, blockSize, assembly='hg19'):
# process coordinate
range1 = mygenome.locus(range1)
range2 = mygenome.locus(range2)
# fetch sequence
seq1 = range1.nibFrag('/data1/Sequence/ucsc_%s' % assembly, blockSize, 0)
seq2 = range2.nibFrag('/data1/Sequence/ucsc_%s' % assembly, 0, blockSize)
# generate block combination
outFile = open('/EQL2/RNASeq_LymphNK/bpTemplate.fa', 'w')
for i in range(len(seq1) - blockSize + 1):
for j in range(len(seq2) - blockSize + 1):
outFile.write('>seq1:%s-%s|seq2:%s-%s\n%s%s\n' %
(i, i + blockSize, j, j + blockSize,
seq1[i:i + blockSize], seq2[j:j + blockSize]))
outFile.close()
# run bowtie
print 'building bowtie index'
os.system(
'bowtie-build -q -f /EQL2/RNASeq_LymphNK/bpTemplate.fa /data1/Sequence/bowtie/bpTemplate'
)
def __init__(self,segL): # take newline- and tab-delimited list ''' __init__ ''' self.segL = segL self.locusL = [mygenome.locus(s[2]) for s in segL]
def main(inGsnapFileName,outReportFileName,sampN,geneNL=[],overlap=10):
eiH, ei_keyH, juncInfoH, ei_cntH = loadAnnot(geneNL)
print 'Finished loading refFlat'
result = mygsnap.gsnapFile(inGsnapFileName,False)
count = 0
for r in result:
if r.nLoci != 1:
continue
match = r.matchL()[0]
for seg in match.segL:
loc = mygenome.locus(seg[2])
if loc.chrSta + overlap > loc.chrEnd - overlap:
continue
cnt_s = findCut(ei_cntH[loc.chrom], ei_keyH[loc.chrom], loc.chrSta + overlap - 1)
cnt_e = findCut(ei_cntH[loc.chrom], ei_keyH[loc.chrom], loc.chrEnd - overlap)
if cnt_e < 1: ## no junction overlaps
continue
elif cnt_s != cnt_e: # overlapping junction exists
pos_min = bisect.bisect_right(ei_keyH[loc.chrom], loc.chrSta + overlap - 1) - 1
pos_max = bisect.bisect_right(ei_keyH[loc.chrom], loc.chrEnd - overlap)
for pos in range(pos_min, pos_max):
if loc.chrSta+overlap <= ei_keyH[loc.chrom][pos] <= loc.chrEnd-overlap:
eiH[loc.chrom][ei_keyH[loc.chrom][pos]] += 1
# count += 1
#
# if count % 10000 == 0:
# print count
outReportFile = open(outReportFileName,'w')
for chrom in ei_keyH.keys():
for e in ei_keyH[chrom]:
if eiH[chrom][e]==[]:
continue
outReportFile.write('%s\t%s\t%s\t%s\n' % (sampN, '%s:%s' % (chrom,e), ','.join(juncInfoH[chrom][e]), eiH[chrom][e]))
def process_bp(inFileName,outFileName,regionL):
result = mygsnap.gsnapFile(inFileName,True)
outFile = open(outFileName, 'w')
outFile.write('browser full knownGene\n')
outFile.write('track name="%s" visibility=2\n' % inFileName)
for rL in result:
if not (rL[0].nLoci==1 and rL[1].nLoci==1 and rL[0].pairRel=='concordant'):
raise Exception
if int(rL[0].pairInfo()[0]) < 76:
continue
locL = [mygenome.locus(rL[0].matchL()[0].segL[0][2]), mygenome.locus(rL[1].matchL()[0].segL[0][2])]
flag = False
for loc in locL:
for region in regionL:
if loc.overlap(region) > 0:
flag = True
if flag:
# print '^%s.*%s$\n' % (rL[0].seq(),mybasic.rc(rL[1].seq())),
# for loc in locL:
# outFile.write('%s\t%s\t%s\n' % (loc.chrom,loc.chrSta,loc.chrEnd))
if locL[0].chrEnd < locL[1].chrSta:
outFile.write('%s\t%s\t%s\n' % (loc.chrom,locL[0].chrEnd,locL[1].chrSta))
else:
outFile.write('%s\t%s\t%s\n' % (loc.chrom,locL[1].chrEnd,locL[0].chrSta))
def main(inGsnapFileName,outReportFileName,sampN,geneNL=[],overlap=10):
eiH, ei_keyH, juncInfoH = loadAnnot(geneNL)
print 'Finished loading refFlat'
result = mygsnap.gsnapFile(inGsnapFileName,False)
count = 0
for r in result:
if r.nLoci != 1:
continue
match = r.matchL()[0]
for seg in match.segL:
loc = mygenome.locus(seg[2])
cursor.execute('select 1 from temp_table where chrom="%s" and pos>=%s and pos<=%s' % (loc.chrom,loc.chrSta+overlap,loc.chrEnd-overlap))
if cursor.fetchone():
eiH[loc.chrom][pos] += 1
count += 1
if count % 10000 == 0:
print count
outReportFile = open(outReportFileName,'w')
for chrom in ei_keyH.keys():
for e in ei_keyH[chrom]:
if eiH[chrom][e]==[]:
continue
outReportFile.write('%s\t%s\t%s\t%s\n' % (sampN, '%s:%s' % (chrom,e), ','.join(juncInfoH[chrom][e]), eiH[chrom][e]))
def genKgCompositeModel(outTextFileName, outFaFileName):
kgH = mygenome.loadKgByChr()
outTextFile = open(outTextFileName, 'w')
outFaFile = open(outFaFileName, 'w')
for chrNum in range(1, 23) + ['X', 'Y', 'M']:
#for chrNum in [1]:
chrom = 'chr%s' % chrNum
txnLocusL_combined = []
for strand in ['+', '-']:
txnLocusL = [
mygenome.locus(
'%s:%s-%s%s' % (chrom, h['txnSta'], h['txnEnd'], strand),
h['kgId'])
for h in filter(lambda x: x['strand'] == strand, kgH[chrom])
]
n_before = len(txnLocusL)
txnLocusL = mygenome.mergeLoci(txnLocusL)
n_after = len(txnLocusL)
#print chrom, strand, n_before, n_after
txnLocusL_combined += txnLocusL
txnLocusL_combined.sort(lambda x, y: cmp(x.chrEnd, y.chrEnd))
txnLocusL_combined.sort(lambda x, y: cmp(x.chrSta, y.chrSta))
for txnLoc in txnLocusL_combined:
exnLocusL = []
for h in filter(lambda x: x['kgId'] in txnLoc.id, kgH[chrom]):
for (exnSta, exnEnd) in h['exnList']:
exnLocusL.append(
mygenome.locus('%s:%s-%s%s' %
(chrom, exnSta, exnEnd, h['strand'])))
exnLocusL.sort(lambda x, y: cmp(x.chrEnd, y.chrEnd))
exnLocusL.sort(lambda x, y: cmp(x.chrSta, y.chrSta))
exnLocusL = mygenome.mergeLoci(exnLocusL)
exnStaL = [str(exnLoc.chrSta) for exnLoc in exnLocusL]
exnEndL = [str(exnLoc.chrEnd) for exnLoc in exnLocusL]
outTextFile.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n' %
(txnLoc.id, txnLoc.chrom, txnLoc.strand,
txnLoc.chrSta, txnLoc.chrEnd, len(exnLocusL),
','.join(exnStaL), ','.join(exnEndL)))
outFaFile.write('>%s|%s|%s|%s|%s\n' %
(txnLoc.id, txnLoc.chrom, txnLoc.strand,
txnLoc.chrSta, txnLoc.chrEnd))
# for exnLoc in exnLocusL:
# outFaFile.write(exnLoc.nibFrag())
txnLocCopy = copy.deepcopy(
txnLoc) # print whole txn sequence in positive strand
txnLocCopy.strand = '+'
outFaFile.write(txnLocCopy.nibFrag())
outFaFile.write('\n')
outTextFile.close()
outFaFile.close()
def exonSkip_proc(inGsnapFileName, outGsnapFileName, outReportFileName, sampN):
geneNameH = mygenome.geneNameH()
geneSetH = mygenome.geneSetH()
geneInfoH = mygenome.geneInfoH(geneNameH, geneSetH)
refFlatH = mygenome.loadRefFlatByChr()
result = mygsnap.gsnapFile(inGsnapFileName, False)
juncHH = {}
for r in result:
match = r.matchL()[0]
if not '(transloc)' in r.pairRel:
raise Exception
if len(match.segL) != 2:
raise Exception
splice_type = re.search('splice_type:([^,\t]*)',
match.segL[0][3]).group(1)
direction = re.search('dir:([^,\t]*)', match.segL[0][3]).group(1)
offset = int(re.search('\.\.([0-9]*)', match.segL[0][1]).group(1))
transcript1 = re.search('label_[12]:([^,\t]*)', match.segL[0][3])
gene1 = set()
if transcript1:
transcript1 = tuple(
[x.split('.exon')[0] for x in transcript1.group(1).split('|')])
for t in transcript1:
g = mygenome.gene(t, geneNameH, geneSetH, geneInfoH)
if g.geneName:
gene1.add(g.geneName)
else:
transcript1 = ()
transcript2 = re.search('label_[12]:([^,\t]*)', match.segL[1][3])
gene2 = set()
if transcript2:
transcript2 = tuple(
[x.split('.exon')[0] for x in transcript2.group(1).split('|')])
for t in transcript2:
g = mygenome.gene(t, geneNameH, geneSetH, geneInfoH)
if g.geneName:
gene2.add(g.geneName)
else:
transcript2 = ()
s1 = match.segL[0][2]
s2 = match.segL[1][2]
bp1 = re.match('([+-])([^:]+):[0-9]+..([0-9]+)', s1)
bp2 = re.match('([+-])([^:]+):([0-9]+)..[0-9]+', s2)
if (bp1.group(1), direction) in (('+', 'sense'), ('-', 'antisense')):
trans_strand1 = '+'
elif (bp1.group(1), direction) in (('+', 'antisense'), ('-', 'sense')):
trans_strand1 = '-'
else:
raise Exception
if (bp2.group(1), direction) in (('+', 'sense'), ('-', 'antisense')):
trans_strand2 = '+'
elif (bp2.group(1), direction) in (('+', 'antisense'), ('-', 'sense')):
trans_strand2 = '-'
else:
raise Exception
bp_gene1 = mygenome.locus(
'%s:%s-%s%s' % (bp1.group(2), int(bp1.group(3)) - 1, bp1.group(3),
trans_strand1)).overlappingGeneL(
refFlatH=refFlatH, strand_sensitive=True)
bp_gene2 = mygenome.locus(
'%s:%s-%s%s' % (bp2.group(2), int(bp2.group(3)) - 1, bp2.group(3),
trans_strand2)).overlappingGeneL(
refFlatH=refFlatH, strand_sensitive=True)
if direction == 'sense':
key = (bp1.groups()[1:], bp2.groups()[1:])
transcript = (transcript1, transcript2)
gene = (tuple(gene1), tuple(gene2))
bp_gene = (bp_gene1, bp_gene2)
elif direction == 'antisense':
key = (bp2.groups()[1:], bp1.groups()[1:])
transcript = (transcript2, transcript1)
gene = (tuple(gene2), tuple(gene1))
bp_gene = (bp_gene2, bp_gene1)
else:
raise Exception
if key in juncHH:
juncHH[key]['match'].append(r)
juncHH[key]['seq'].append(r.seq())
juncHH[key]['reg'].append((direction, offset))
else:
juncHH[key] = {
'match': [r],
'splice_type': splice_type,
'seq': [r.seq()],
'reg': [(direction, offset)],
'transcript': transcript,
'gene': gene,
'bp_gene': bp_gene
}
juncKH = juncHH.items()
juncKH.sort(lambda x, y: cmp(len(set(y[1]['reg'])), len(set(x[1]['reg']))))
outGsnapFile = open(outGsnapFileName, 'w')
outReportFile = open(outReportFileName, 'w')
for (key, juncH) in juncKH:
if key[0][0] == key[1][0]:
type = 'intra'
else:
type = 'inter'
geneInfo1 = []
censusInfo1 = []
for geneName in juncH['gene'][0]:
gene = mygenome.gene(geneName, geneNameH, geneSetH, geneInfoH)
geneInfo1.append(
'%s:%s:%s' %
(geneName, gene.getAttr('desc'), gene.getAttr('summary')))
censusInfo1.append('%s:%s:%s:%s' %
(gene.getAttr('census_somatic'),
gene.getAttr('census_germline'),
gene.getAttr('census_mutType'),
gene.getAttr('census_translocPartners')))
geneInfo2 = []
censusInfo2 = []
for geneName in juncH['gene'][1]:
gene = mygenome.gene(geneName, geneNameH, geneSetH, geneInfoH)
geneInfo2.append(
'%s:%s:%s' %
(geneName, gene.getAttr('desc'), gene.getAttr('summary')))
censusInfo2.append('%s:%s:%s:%s' %
(gene.getAttr('census_somatic'),
gene.getAttr('census_germline'),
gene.getAttr('census_mutType'),
gene.getAttr('census_translocPartners')))
outReportFile.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n' % \
(type, juncH['splice_type'], sampN, ':'.join(key[0]), ':'.join(key[1]), \
';'.join(juncH['transcript'][0]), ';'.join(juncH['transcript'][1]), ';'.join(juncH['gene'][0]), ';'.join(juncH['gene'][1]), ';'.join(geneInfo1), ';'.join(geneInfo2), \
';'.join(censusInfo1), ';'.join(censusInfo2), ','.join(juncH['bp_gene'][0]), ','.join(juncH['bp_gene'][1]), \
len(juncH['match']) ,len(set(juncH['seq'])), len(set(juncH['reg']))))
for m in juncH['match']:
outGsnapFile.write(m.rawText() + '\n')
def genCompositeModel(outTextFileName,outFaFileName,intronSize=100):
geneNameH = mygenome.geneNameH()
geneSetH = mygenome.geneSetH()
geneInfoH = mygenome.geneInfoH(geneNameH,geneSetH)
geneH = mygenome.loadKgByChr()
#geneH = mygenome.loadLincByChr(h=geneH)
outTextFile = open(outTextFileName, 'w')
outFaFile = open(outFaFileName, 'w')
for chrNum in range(1,23)+['X','Y','M']:
#for chrNum in [1]:
chrom = 'chr%s' % chrNum
geneH_byChr = filter(lambda x: mygenome.gene(x['geneId'],geneNameH,geneSetH,geneInfoH).geneName in mygenome.TK, geneH[chrom])
txnLocusL_combined = []
for strand in ['+','-']:
txnLocusL = [mygenome.locus('%s:%s-%s%s' % (chrom,h['txnSta'],h['txnEnd'],strand),h['geneId']) for h in filter(lambda x: x['strand']==strand, geneH_byChr)]
n_before = len(txnLocusL)
txnLocusL = mygenome.mergeLoci(txnLocusL)
n_after = len(txnLocusL)
#print chrom, strand, n_before, n_after
txnLocusL_combined += txnLocusL
txnLocusL_combined.sort(lambda x,y: cmp(x.chrEnd,y.chrEnd))
txnLocusL_combined.sort(lambda x,y: cmp(x.chrSta,y.chrSta))
for txnLoc in txnLocusL_combined:
exnLocusL = []
for h in filter(lambda x: x['geneId'] in txnLoc.id, geneH_byChr):
for (exnSta,exnEnd) in h['exnList']:
exnLocusL.append(mygenome.locus('%s:%s-%s%s' % (chrom, exnSta, exnEnd, h['strand'])))
exnLocusL.sort(lambda x,y: cmp(x.chrEnd,y.chrEnd))
exnLocusL.sort(lambda x,y: cmp(x.chrSta,y.chrSta))
exnLocusL = mygenome.mergeLoci(exnLocusL)
exnStaL = [str(exnLoc.chrSta) for exnLoc in exnLocusL]
exnEndL = [str(exnLoc.chrEnd) for exnLoc in exnLocusL]
outTextFile.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n' % (txnLoc.id,txnLoc.chrom,txnLoc.strand,txnLoc.chrSta,txnLoc.chrEnd,len(exnLocusL),','.join(exnStaL),','.join(exnEndL)))
outFaFile.write('>%s|%s|%s|%s|%s\n' % (txnLoc.id,txnLoc.chrom,txnLoc.strand,txnLoc.chrSta,txnLoc.chrEnd))
for i in range(len(exnLocusL)):
exnLocCopy = copy.deepcopy(exnLocusL[i])
exnLocCopy.strand = '+'
if i > 0:
exnLocCopy.chrSta -= min(intronSize, int((exnLocusL[i].chrSta - exnLocusL[i-1].chrEnd)/2))
if i < len(exnLocusL)-1:
exnLocCopy.chrEnd += min(intronSize, int((exnLocusL[i+1].chrSta - exnLocusL[i].chrEnd)/2))
outFaFile.write(exnLocCopy.nibFrag())
outFaFile.write('\n')
outTextFile.close()
outFaFile.close()
def gsnap_process_junction(inGsnapFileName,outGsnapFileName,outReportFileName,sampN):
geneNameH = mygenome.geneNameH()
geneSetH = mygenome.geneSetH()
geneInfoH = mygenome.geneInfoH(geneNameH,geneSetH)
refFlatH = mygenome.loadRefFlatByChr()
result = mygsnap.gsnapFile(inGsnapFileName,False)
juncHH = {}
for r in result:
match = r.matchL()[0]
if not '(transloc)' in r.pairRel:
raise Exception
if len(match.segL) != 2:
raise Exception
splice_type = re.search('splice_type:([^,\t]*)', match.segL[0][3]).group(1)
direction = re.search('dir:([^,\t]*)', match.segL[0][3]).group(1)
offset = int(re.search('\.\.([0-9]*)', match.segL[0][1]).group(1))
rm = re.search('label_[12]:([^,\t]*)', match.segL[0][3])
gene1 = set()
if rm:
trans_exon1 = rm.group(1).split('|')
for t in trans_exon1:
g = mygenome.gene(t.split('.exon')[0],geneNameH,geneSetH,geneInfoH)
if g.geneName:
gene1.add(g.geneName)
else:
trans_exon1 = ()
rm = re.search('label_[12]:([^,\t]*)', match.segL[0][3])
gene2 = set()
if rm:
trans_exon2 = rm.group(1).split('|')
for t in trans_exon2:
g = mygenome.gene(t.split('.exon')[0],geneNameH,geneSetH,geneInfoH)
if g.geneName:
gene2.add(g.geneName)
else:
trans_exon2 = ()
s1 = match.segL[0][2]
s2 = match.segL[1][2]
bp1 = re.match('([+-])([^:]+):[0-9]+..([0-9]+)',s1)
bp2 = re.match('([+-])([^:]+):([0-9]+)..[0-9]+',s2)
if (bp1.group(1),direction) in (('+','sense'),('-','antisense')):
trans_strand1 = '+'
elif (bp1.group(1),direction) in (('+','antisense'),('-','sense')):
trans_strand1 = '-'
else:
raise Exception
if (bp2.group(1),direction) in (('+','sense'),('-','antisense')):
trans_strand2 = '+'
elif (bp2.group(1),direction) in (('+','antisense'),('-','sense')):
trans_strand2 = '-'
else:
raise Exception
locus1 = mygenome.locus('%s:%s-%s%s' % (bp1.group(2),int(bp1.group(3))-1,bp1.group(3),trans_strand1))
bp_gene1 = list(set(locus1.overlappingGeneL(refFlatH=refFlatH,strand_sensitive=True)).difference(gene1))
locus2 = mygenome.locus('%s:%s-%s%s' % (bp2.group(2),int(bp2.group(3))-2,bp2.group(3),trans_strand2))
bp_gene2 = list(set(locus2.overlappingGeneL(refFlatH=refFlatH,strand_sensitive=True)).difference(gene2))
if direction=='sense':
key = (bp1.groups()[1:],bp2.groups()[1:])
trans_exon = (trans_exon1,trans_exon2)
gene = (list(gene1),list(gene2))
bp_gene = (bp_gene1,bp_gene2)
elif direction=='antisense':
key = (bp2.groups()[1:],bp1.groups()[1:])
trans_exon = (trans_exon2,trans_exon1)
gene = (list(gene2),list(gene1))
bp_gene = (bp_gene2,bp_gene1)
else:
raise Exception
if key in juncHH:
juncHH[key]['match'].append(r)
juncHH[key]['seq'].append(r.seq())
juncHH[key]['reg'].append((direction,offset))
else:
juncHH[key] = {'match':[r], 'splice_type':splice_type, 'seq':[r.seq()], 'reg':[(direction,offset)], 'trans_exon':trans_exon, 'gene':gene, 'bp_gene':bp_gene}
juncKH = juncHH.items()
juncKH.sort(lambda x,y: cmp(len(set(y[1]['reg'])),len(set(x[1]['reg']))))
outGsnapFile = open(outGsnapFileName,'w')
outReportFile = open(outReportFileName,'w')
for (key, juncH) in juncKH:
if key[0][0] == key[1][0]:
type = 'intra'
else:
type = 'inter'
geneInfo1 = []
censusInfo1 = []
for geneName in juncH['gene'][0]+juncH['bp_gene'][0]:
gene = mygenome.gene(geneName,geneNameH,geneSetH,geneInfoH)
geneInfo1.append('%s:%s:%s' % (geneName,gene.getAttr('desc'),gene.getAttr('summary')))
censusInfo1.append('%s:%s:%s:%s' % (gene.getAttr('census_somatic'),gene.getAttr('census_germline'),gene.getAttr('census_mutType'),gene.getAttr('census_translocPartners')))
geneInfo2 = []
censusInfo2 = []
for geneName in juncH['gene'][1]+juncH['bp_gene'][1]:
gene = mygenome.gene(geneName,geneNameH,geneSetH,geneInfoH)
geneInfo2.append('%s:%s:%s' % (geneName,gene.getAttr('desc'),gene.getAttr('summary')))
censusInfo2.append('%s:%s:%s:%s' % (gene.getAttr('census_somatic'),gene.getAttr('census_germline'),gene.getAttr('census_mutType'),gene.getAttr('census_translocPartners')))
outReportFile.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s;%s\t%s;%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n' % \
(type, juncH['splice_type'], sampN, ':'.join(key[0]), ':'.join(key[1]), \
','.join(juncH['trans_exon'][0]), ','.join(juncH['trans_exon'][1]), \
','.join(juncH['gene'][0]), ','.join(juncH['bp_gene'][0]), ','.join(juncH['gene'][1]), ','.join(juncH['bp_gene'][1]), \
';'.join(geneInfo1), ';'.join(geneInfo2), ';'.join(censusInfo1), ';'.join(censusInfo2), \
len(juncH['match']) ,len(set(juncH['seq'])), len(set(juncH['reg']))))
for m in juncH['match']:
outGsnapFile.write(m.rawText()+'\n')
def fusion_proc_annot(inReportFileName,
outReportFileName,
inCnaGctFileName=None):
geneDB = mygenome.getGeneDB()
frameInfoH = mygenome.getFrameInfoH()
refFlatH = mygenome.loadRefFlatByChr()
if inCnaGctFileName:
cnaDB = mygenome.tcgaCnaDB(inCnaGctFileName)
else:
cnaDB = None
outReportFile = open(outReportFileName, 'w')
for line in open(inReportFileName):
(splice_type, sampN, bp1, bp2, teStr1, teStr2, nmatch, nseq,
nreg) = line[:-1].split('\t')
if inCnaGctFileName:
indivId = re.match('.*(TCGA-[0-9]{2}-[0-9]{4}).*', sampN).group(1)
geneStatL = []
for (bp, teStr) in ((bp1, teStr1), (bp2, teStr2)):
geneS = set()
teL = []
for te in teStr.split(','):
rm = re.match('(.*)\.exon([0-9]*)/[0-9]*', te)
if rm:
t = rm.group(1)
e = int(rm.group(2))
g = mygenome.gene(t, geneDB=geneDB)
if g.geneName:
geneS.add(g.geneName)
teL.append((t, e))
rm = re.match('([+-])(chr[^:]*):([0-9]*)', bp)
bp_geneS = set(
mygenome.locus('%s:%s-%s%s' %
(rm.group(2), int(rm.group(3)) - 1, rm.group(3),
rm.group(1))).overlappingGeneL(
refFlatH=refFlatH, strand_sensitive=True))
bp_geneS = bp_geneS.difference(geneS)
cnaInfo = []
geneInfo = []
censusInfo = []
goInfoS = set()
keggInfoS = set()
biocartaInfoS = set()
for geneName in list(geneS) + list(bp_geneS):
gene = mygenome.gene(geneName, geneDB=geneDB)
if cnaDB:
cnaInfo.append('%s:%s' %
(geneName, cnaDB.query(indivId, geneName)))
geneInfo.append(
'%s:%s:%s' %
(geneName, gene.getAttr('desc'), gene.getAttr('summary')))
censusInfo.append('%s:%s:%s:%s' %
(gene.getAttr('census_somatic'),
gene.getAttr('census_germline'),
gene.getAttr('census_mutType'),
gene.getAttr('census_translocPartners')))
goInfoS = goInfoS.union(set(gene.getAttr('go')))
keggInfoS = keggInfoS.union(set(gene.getAttr('kegg')))
biocartaInfoS = biocartaInfoS.union(
set(gene.getAttr('biocarta')))
geneStatL.append(
(bp1.split(':')[0], bp, teStr, teL, geneS, bp_geneS, cnaInfo,
geneInfo, censusInfo, goInfoS, keggInfoS, biocartaInfoS))
(chrom1, bp1, teStr1, teL1, geneS1, bp_geneS1, cnaInfo1, geneInfo1,
censusInfo1, goInfoS1, keggInfoS1, biocartaInfoS1) = geneStatL[0]
(chrom2, bp2, teStr2, teL2, geneS2, bp_geneS2, cnaInfo2, geneInfo2,
censusInfo2, goInfoS2, keggInfoS2, biocartaInfoS2) = geneStatL[1]
if chrom1 == chrom2:
type = 'intra'
else:
type = 'inter'
frameL = []
for (t1, e1) in teL1:
for (t2, e2) in teL2:
cons = mygenome.frameCons(t1, e1, t2, e2, frameInfoH)
if cons == 'Y':
frameL.append('%s.%s-%s.%s:%s' % (t1, e1, t2, e2, cons))
outReportFile.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n' % \
(sampN, splice_type, type, bp1,bp2, teStr1,teStr2, ','.join(frameL), ','.join(cnaInfo1), ','.join(cnaInfo2), \
'%s;%s' % (','.join(geneS1),','.join(bp_geneS1)), ';'.join(geneInfo1), ';'.join(censusInfo1), \
';'.join(map(str,goInfoS1)), ';'.join(map(str,keggInfoS1)), ';'.join(map(str,biocartaInfoS1)),
'%s;%s' % (','.join(geneS2),','.join(bp_geneS2)), ';'.join(geneInfo2), ';'.join(censusInfo2), \
';'.join(map(str,goInfoS2)), ';'.join(map(str,keggInfoS2)), ';'.join(map(str,biocartaInfoS2)),
nmatch,nseq,nreg))
def genKgCompositeModel(outTextFileName,outFaFileName):
kgH = mygenome.loadKgByChr()
outTextFile = open(outTextFileName, 'w')
outFaFile = open(outFaFileName, 'w')
for chrNum in range(1,23)+['X','Y','M']:
#for chrNum in [1]:
chrom = 'chr%s' % chrNum
txnLocusL_combined = []
for strand in ['+','-']:
txnLocusL = [mygenome.locus('%s:%s-%s%s' % (chrom,h['txnSta'],h['txnEnd'],strand),h['kgId']) for h in filter(lambda x: x['strand']==strand, kgH[chrom])]
n_before = len(txnLocusL)
txnLocusL = mygenome.mergeLoci(txnLocusL)
n_after = len(txnLocusL)
#print chrom, strand, n_before, n_after
txnLocusL_combined += txnLocusL
txnLocusL_combined.sort(lambda x,y: cmp(x.chrEnd,y.chrEnd))
txnLocusL_combined.sort(lambda x,y: cmp(x.chrSta,y.chrSta))
for txnLoc in txnLocusL_combined:
exnLocusL = []
for h in filter(lambda x: x['kgId'] in txnLoc.id, kgH[chrom]):
for (exnSta,exnEnd) in h['exnList']:
exnLocusL.append(mygenome.locus('%s:%s-%s%s' % (chrom, exnSta, exnEnd, h['strand'])))
exnLocusL.sort(lambda x,y: cmp(x.chrEnd,y.chrEnd))
exnLocusL.sort(lambda x,y: cmp(x.chrSta,y.chrSta))
exnLocusL = mygenome.mergeLoci(exnLocusL)
exnStaL = [str(exnLoc.chrSta) for exnLoc in exnLocusL]
exnEndL = [str(exnLoc.chrEnd) for exnLoc in exnLocusL]
outTextFile.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n' % (txnLoc.id,txnLoc.chrom,txnLoc.strand,txnLoc.chrSta,txnLoc.chrEnd,len(exnLocusL),','.join(exnStaL),','.join(exnEndL)))
outFaFile.write('>%s|%s|%s|%s|%s\n' % (txnLoc.id,txnLoc.chrom,txnLoc.strand,txnLoc.chrSta,txnLoc.chrEnd))
# for exnLoc in exnLocusL:
# outFaFile.write(exnLoc.nibFrag())
txnLocCopy = copy.deepcopy(txnLoc) # print whole txn sequence in positive strand
txnLocCopy.strand = '+'
outFaFile.write(txnLocCopy.nibFrag())
outFaFile.write('\n')
outTextFile.close()
outFaFile.close()
def fusion_proc_annot(inReportFileName,outReportFileName,inCnaGctFileName=None):
geneDB = mygenome.getGeneDB()
frameInfoH = mygenome.getFrameInfoH()
refFlatH = mygenome.loadRefFlatByChr()
if inCnaGctFileName:
cnaDB = mygenome.tcgaCnaDB(inCnaGctFileName)
else:
cnaDB = None
outReportFile = open(outReportFileName,'w')
for line in open(inReportFileName):
(splice_type,sampN,bp1,bp2,teStr1,teStr2,nmatch,nseq,nreg) = line[:-1].split('\t')
if inCnaGctFileName:
indivId = re.match('.*(TCGA-[0-9]{2}-[0-9]{4}).*',sampN).group(1)
geneStatL = []
for (bp,teStr) in ((bp1,teStr1),(bp2,teStr2)):
geneS = set()
teL = []
for te in teStr.split(','):
rm = re.match('(.*)\.exon([0-9]*)/[0-9]*',te)
if rm:
t = rm.group(1)
e = int(rm.group(2))
g = mygenome.gene(t,geneDB=geneDB)
if g.geneName:
geneS.add(g.geneName)
teL.append((t,e))
rm = re.match('([+-])(chr[^:]*):([0-9]*)',bp)
bp_geneS = set(mygenome.locus('%s:%s-%s%s' % (rm.group(2),int(rm.group(3))-1,rm.group(3),rm.group(1))).overlappingGeneL(refFlatH=refFlatH,strand_sensitive=True))
bp_geneS = bp_geneS.difference(geneS)
cnaInfo = []
geneInfo = []
censusInfo = []
goInfoS = set()
keggInfoS = set()
biocartaInfoS = set()
for geneName in list(geneS) + list(bp_geneS):
gene = mygenome.gene(geneName,geneDB=geneDB)
if cnaDB:
cnaInfo.append('%s:%s' % (geneName,cnaDB.query(indivId,geneName)))
geneInfo.append('%s:%s:%s' % (geneName,gene.getAttr('desc'),gene.getAttr('summary')))
censusInfo.append('%s:%s:%s:%s' % (gene.getAttr('census_somatic'),gene.getAttr('census_germline'),gene.getAttr('census_mutType'),gene.getAttr('census_translocPartners')))
goInfoS = goInfoS.union(set(gene.getAttr('go')))
keggInfoS = keggInfoS.union(set(gene.getAttr('kegg')))
biocartaInfoS = biocartaInfoS.union(set(gene.getAttr('biocarta')))
geneStatL.append((bp1.split(':')[0],bp,teStr,teL,geneS,bp_geneS,cnaInfo,geneInfo,censusInfo,goInfoS,keggInfoS,biocartaInfoS))
(chrom1,bp1,teStr1,teL1,geneS1,bp_geneS1,cnaInfo1,geneInfo1,censusInfo1,goInfoS1,keggInfoS1,biocartaInfoS1) = geneStatL[0]
(chrom2,bp2,teStr2,teL2,geneS2,bp_geneS2,cnaInfo2,geneInfo2,censusInfo2,goInfoS2,keggInfoS2,biocartaInfoS2) = geneStatL[1]
if chrom1 == chrom2:
type = 'intra'
else:
type = 'inter'
frameL = []
for (t1,e1) in teL1:
for (t2,e2) in teL2:
cons = mygenome.frameCons(t1,e1, t2,e2, frameInfoH)
if cons=='Y':
frameL.append('%s.%s-%s.%s:%s' % (t1,e1,t2,e2,cons))
outReportFile.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\n' % \
(sampN, splice_type, type, bp1,bp2, teStr1,teStr2, ','.join(frameL), ','.join(cnaInfo1), ','.join(cnaInfo2), \
'%s;%s' % (','.join(geneS1),','.join(bp_geneS1)), ';'.join(geneInfo1), ';'.join(censusInfo1), \
';'.join(map(str,goInfoS1)), ';'.join(map(str,keggInfoS1)), ';'.join(map(str,biocartaInfoS1)),
'%s;%s' % (','.join(geneS2),','.join(bp_geneS2)), ';'.join(geneInfo2), ';'.join(censusInfo2), \
';'.join(map(str,goInfoS2)), ';'.join(map(str,keggInfoS2)), ';'.join(map(str,biocartaInfoS2)),
nmatch,nseq,nreg))
