def runBAMtoJunctionBED(paths_to_run):
bamfile_dir,bed_reference_dir,output_bedfile_path = paths_to_run
output_bedfile_path = string.replace(bamfile_dir,'.bam','__junction.bed')
#if os.path.exists(output_bedfile_path) == False: ### Only run if the file doesn't exist
results = BAMtoJunctionBED.parseJunctionEntries(bamfile_dir,multi=True)
#else: print output_bedfile_path, 'already exists.'
return results
def runBAMtoJunctionBED(paths_to_run):
bamfile_dir, bed_reference_dir, output_bedfile_path = paths_to_run
output_bedfile_path = string.replace(bamfile_dir, '.bam', '__junction.bed')
#if os.path.exists(output_bedfile_path) == False: ### Only run if the file doesn't exist
results = BAMtoJunctionBED.parseJunctionEntries(bamfile_dir, multi=True)
#else: print output_bedfile_path, 'already exists.'
return results
def parseExonReferences(bam_dir,reference_exon_bed,multi=False):
start_time = time.time()
bamfile = pysam.Samfile(bam_dir, "rb" )
reference_rows=0
output_bed_rows=0
o = open (string.replace(bam_dir,'.bam','__exon.bed'),"w")
for line in open(reference_exon_bed,'rU').xreadlines(): ### read each line one-at-a-time rather than loading all in memory
line = line.rstrip('\n')
reference_rows+=1
ref_entries = string.split(line,'\t'); #'12', '6998470', '6998522', 'ENSG00000111671:E1.1_ENSE00001754003', '0', '-'
chr,start,stop,exon,null,strand = ref_entries[:6]
read_count=0
try:
#if exon == 'ENSMUSG00000001472:E17.1':
#chr = '12'; start = '6998470'; stop = '6998522'
for alignedread in bamfile.fetch(chr, int(start),int(stop)):
proceed = True
try: cigarstring = alignedread.cigarstring
except Exception:
codes = map(lambda x: x[0],alignedread.cigar)
if 3 in codes: cigarstring = 'N'
else: cigarstring = None
if cigarstring == None: pass
else:
### Exclude junction reads ("N")
if 'N' in cigarstring:
X=int(alignedread.pos)
Y=int(alignedread.pos+alignedread.alen)
start= int(start)
stop = int(stop)
proceed = False
a = [X,Y]; a.sort()
b = [X,Y,start,stop]; b.sort()
if a[0]==b[1] or a[1]==b[2]: ### Hence, the read starts or ends in that interval
proceed = True
if proceed == False:
### Also search for cases were part of the read is contained within the exon
import BAMtoJunctionBED
coordinates,up_to_intron_dist = BAMtoJunctionBED.getSpliceSites(alignedread.cigar,X)
for (five_prime_ss,three_prime_ss) in coordinates:
five_prime_ss,three_prime_ss=int(five_prime_ss),int(three_prime_ss)
if five_prime_ss==start or three_prime_ss==start or five_prime_ss==stop or three_prime_ss==stop:
proceed = True
#print five_prime_ss, three_prime_ss, start, stop;sys.exit()
if proceed: read_count+=1
entries = [chr,str(start),str(stop),exon,null,strand,str(read_count),'0',str(int(stop)-int(start)),'0']
o.write(string.join(entries,'\t')+'\n')
output_bed_rows+=1
except Exception,e:
### Occurs also due to non-chromosome contigs in the annotation file
if 'bamfile without index' in e:
print 'Please ensure an index exists for the bam file:',bam_dir;sys.exit()
def parseExonReferences(bam_dir,reference_exon_bed,multi=False):
start_time = time.time()
bamfile = pysam.Samfile(bam_dir, "rb" )
reference_rows=0
output_bed_rows=0
o = open (string.replace(bam_dir,'.bam','__exon.bed'),"w")
io = AppendOrWrite(string.replace(bam_dir,'.bam','__junction.bed'))
for line in open(reference_exon_bed,'rU').xreadlines(): ### read each line one-at-a-time rather than loading all in memory
line = line.rstrip('\n')
reference_rows+=1
ref_entries = string.split(line,'\t'); #'12', '6998470', '6998522', 'ENSG00000111671:E1.1_ENSE00001754003', '0', '-'
chr,start,stop,exon,null,strand = ref_entries[:6]
read_count=0;
five_intron_junction_count=0
three_intron_junction_count=0
try:
#if exon == 'ENSMUSG00000001472:E17.1':
#chr = '12'; start = '6998470'; stop = '6998522'
for alignedread in bamfile.fetch(chr, int(start),int(stop)):
proceed = True
try: cigarstring = alignedread.cigarstring
except Exception:
codes = map(lambda x: x[0],alignedread.cigar)
if 3 in codes: cigarstring = 'N'
else: cigarstring = None
try: read_strand = alignedread.opt('XS') ### TopHat/STAR knows which sequences are likely real splice sites so it assigns a real strand to the read
except Exception,e:
#if multi == False: print 'No TopHat strand information';sys.exit()
read_strand = None ### TopHat doesn't predict strand for many reads
if read_strand==None or read_strand==strand: ### Tries to ensure the propper strand reads are considered (if strand read info available)
if cigarstring == None: pass
else:
### Exclude junction reads ("N")
if 'N' in cigarstring:
X=int(alignedread.pos)
Y=int(alignedread.pos+alignedread.alen)
start= int(start)
stop = int(stop)
proceed = False
a = [X,Y]; a.sort()
b = [X,Y,start,stop]; b.sort()
if a[0]==b[1] or a[1]==b[2]: ### Hence, the read starts or ends in that interval
proceed = True
if proceed == False:
### Also search for cases were part of the read is contained within the exon
import BAMtoJunctionBED
coordinates,up_to_intron_dist = BAMtoJunctionBED.getSpliceSites(alignedread.cigar,X)
for (five_prime_ss,three_prime_ss) in coordinates:
five_prime_ss,three_prime_ss=int(five_prime_ss),int(three_prime_ss)
if five_prime_ss==start or three_prime_ss==start or five_prime_ss==stop or three_prime_ss==stop:
proceed = True
#print five_prime_ss, three_prime_ss, start, stop;sys.exit()
else:
### Below code is for more accurate estimation of intron retention
try:
if 'I' in exon:
X=int(alignedread.pos)
Y=int(alignedread.pos+alignedread.alen)
start= int(start)
stop = int(stop)
a1 = [X,Y]; a1.sort()
b1 = [X,Y,start,stop]; b1.sort()
if a1[0]==b1[0] or a1[1]==b1[-1]: ### Hence, the read starts or ends OUTSIDE of that interval
mate = bamfile.mate(alignedread) ### looup the paired-end mate for this read
try: cigarstring = mate.cigarstring
except Exception:
codes = map(lambda x: x[0],mate.cigar)
if 3 in codes: cigarstring = 'N'
else: cigarstring = None
if 'N' not in cigarstring:
RX=int(mate.pos)
RY=int(mate.pos+mate.alen)
a2 = [start,stop]; a2.sort()
b2 = [RX,RY,start,stop]; b2.sort()
if a2[0]==b2[0] and a2[-1]==b2[-1]:
if a1[0]==b1[0]:
five_intron_junction_count+=1 ### intron junction read that spans the 5' intron-exon
#print exon, start, stop, X, Y, RX, RY, strand, read_strand;sys.exit()
elif a1[1]==b1[-1]:
three_intron_junction_count+=1 ### intron junction read that spans the 3' intron-exon
except Exception,e: ### Usually an unmapped read
#print traceback.format_exc();sys.exit()
pass
if proceed: read_count+=1
entries = [chr,str(start),str(stop),exon,null,strand,str(read_count),'0',str(int(stop)-int(start)),'0']
o.write(string.join(entries,'\t')+'\n')
output_bed_rows+=1
if 'I' in exon and five_intron_junction_count>0 and three_intron_junction_count>0:
if strand=='-': increment = -1
else: increment = -1
outlier_start = start-10+increment; outlier_end = start+10+increment
junction_id = exon+'-'+str(start)
exon_lengths = '10,10'; dist = '0,0'
entries = [chr,str(outlier_start),str(outlier_end),junction_id,str(five_intron_junction_count),strand,str(outlier_start),str(outlier_end),'255,0,0\t2',exon_lengths,'0,'+dist]
io.write(string.join(entries,'\t')+'\n')
### 3' junction
if strand=='-': increment = 0
else: increment = 0
outlier_start = stop-10+increment; outlier_end = stop+10+increment
junction_id = exon+'-'+str(stop)
exon_lengths = '10,10'; dist = '0,0'
entries = [chr,str(outlier_start),str(outlier_end),junction_id,str(three_intron_junction_count),strand,str(outlier_start),str(outlier_end),'255,0,0\t2',exon_lengths,'0,'+dist]
io.write(string.join(entries,'\t')+'\n')
def parseExonReferences(bam_dir, reference_exon_bed, multi=False):
start_time = time.time()
bamfile = pysam.Samfile(bam_dir, "rb")
reference_rows = 0
output_bed_rows = 0
o = open(string.replace(bam_dir, '.bam', '__exon.bed'), "w")
for line in open(reference_exon_bed, 'rU').xreadlines(
): ### read each line one-at-a-time rather than loading all in memory
line = line.rstrip('\n')
reference_rows += 1
ref_entries = string.split(line, '\t')
#'12', '6998470', '6998522', 'ENSG00000111671:E1.1_ENSE00001754003', '0', '-'
chr, start, stop, exon, null, strand = ref_entries[:6]
read_count = 0
try:
#if exon == 'ENSMUSG00000001472:E17.1':
#chr = '12'; start = '6998470'; stop = '6998522'
for alignedread in bamfile.fetch(chr, int(start), int(stop)):
proceed = True
try:
cigarstring = alignedread.cigarstring
except Exception:
codes = map(lambda x: x[0], alignedread.cigar)
if 3 in codes: cigarstring = 'N'
else: cigarstring = None
if cigarstring == None: pass
else:
### Exclude junction reads ("N")
if 'N' in cigarstring:
X = int(alignedread.pos)
Y = int(alignedread.pos + alignedread.alen)
start = int(start)
stop = int(stop)
proceed = False
a = [X, Y]
a.sort()
b = [X, Y, start, stop]
b.sort()
if a[0] == b[1] or a[1] == b[
2]: ### Hence, the read starts or ends in that interval
proceed = True
if proceed == False:
### Also search for cases were part of the read is contained within the exon
import BAMtoJunctionBED
coordinates, up_to_intron_dist = BAMtoJunctionBED.getSpliceSites(
alignedread.cigar, X)
for (five_prime_ss, three_prime_ss) in coordinates:
five_prime_ss, three_prime_ss = int(
five_prime_ss), int(three_prime_ss)
if five_prime_ss == start or three_prime_ss == start or five_prime_ss == stop or three_prime_ss == stop:
proceed = True
#print five_prime_ss, three_prime_ss, start, stop;sys.exit()
if proceed: read_count += 1
entries = [
chr,
str(start),
str(stop), exon, null, strand,
str(read_count), '0',
str(int(stop) - int(start)), '0'
]
o.write(string.join(entries, '\t') + '\n')
output_bed_rows += 1
except Exception, e:
### Occurs also due to non-chromosome contigs in the annotation file
if 'bamfile without index' in e:
print 'Please ensure an index exists for the bam file:', bam_dir
sys.exit()
