def bam_find_regions(bam_name, merge_distance=10, min_read_count=2, only_uniq_starts=False, nostrand=False, out=sys.stdout):
bamfile = bam_open(bam_name)
region_plus = None
region_minus = None
for pileup in bam_pileup_iter(bamfile, mask=1540):
chrom = bamfile.getrname(pileup.tid)
for read in pileup.pileups:
if read.is_del:
continue
if nostrand or not read.alignment.is_reverse:
if not region_plus or region_plus.chrom != chrom or (region_plus.end + merge_distance) < pileup.pos:
if region_plus and region_plus.read_count >= min_read_count:
region_plus.write(out)
region_plus = ExpressedRegion(chrom, only_uniq_starts)
region_plus.add_column(read, pileup.pos)
else:
if not region_minus or region_minus.chrom != chrom or (region_minus.end + merge_distance) < pileup.pos:
if region_minus and region_minus.read_count >= min_read_count:
region_minus.write(out)
region_minus = ExpressedRegion(chrom, only_uniq_starts)
region_minus.add_column(read, pileup.pos)
if region_plus and region_plus.read_count >= min_read_count:
region_plus.write(out)
if region_minus and region_minus.read_count >= min_read_count:
region_minus.write(out)
bamfile.close()
def bam_cims_finder(bam_fnames, output='bed', ref_fname=None, flanking=12, cutoff=0.1, stranded=True, window_size=20):
for bam_fname in bam_fnames:
sys.stderr.write('%s\n' % bam_fname)
bam = pysam.Samfile(bam_fname, "rb")
if output == 'fasta':
emitter = FASTAEmitter(ref_fname, flanking)
else:
emitter = BEDEmitter()
if stranded:
strands = ['+', '-']
else:
strands = ['']
for strand in strands:
manager = RegionManager(emitter, strand, window_size)
for pileup in bam_pileup_iter(bam, mask=1540):
chrom = bam.getrname(pileup.tid)
deletions = 0.0
total = 0.0
del_reads = set()
total_reads = set()
for pileupread in pileup.pileups:
if not strand or (strand == '+' and not pileupread.alignment.is_reverse) or (strand == '-' and pileupread.alignment.is_reverse):
if is_read_match_at_pos(pileupread.alignment, pileupread.qpos):
total += 1
total_reads.add(pileupread.alignment.qname)
if is_read_del_at_pos(pileupread.alignment, pileupread.qpos):
deletions += 1
del_reads.add(pileupread.alignment.qname)
# print ""
# print chrom
# print pileup.pos
# print pileupread.alignment.qname
# print pileupread.alignment.pos
# print pileupread.alignment.cigar
# print pileupread.qpos
if total > 0:
pct = deletions / total
if pct > cutoff:
manager.add(chrom, pileup.pos, strand, del_reads, total_reads)
manager.close()
bam.close()
emitter.close()
def bam_find_regions(bam_name,
merge_distance=10,
min_read_count=2,
only_uniq_starts=False,
nostrand=False,
out=sys.stdout):
bamfile = bam_open(bam_name)
region_plus = None
region_minus = None
for pileup in bam_pileup_iter(bamfile, mask=1540):
chrom = bamfile.getrname(pileup.tid)
for read in pileup.pileups:
if read.is_del:
continue
if nostrand or not read.alignment.is_reverse:
if not region_plus or region_plus.chrom != chrom or (
region_plus.end + merge_distance) < pileup.pos:
if region_plus and region_plus.read_count >= min_read_count:
region_plus.write(out)
region_plus = ExpressedRegion(chrom, only_uniq_starts)
region_plus.add_column(read, pileup.pos)
else:
if not region_minus or region_minus.chrom != chrom or (
region_minus.end + merge_distance) < pileup.pos:
if region_minus and region_minus.read_count >= min_read_count:
region_minus.write(out)
region_minus = ExpressedRegion(chrom, only_uniq_starts)
region_minus.add_column(read, pileup.pos)
if region_plus and region_plus.read_count >= min_read_count:
region_plus.write(out)
if region_minus and region_minus.read_count >= min_read_count:
region_minus.write(out)
bamfile.close()
def bam_minorallele(bam_fname, ref_fname, min_qual=0, min_count=0, num_alleles=0, name=None, min_ci_low=None):
bam = pysam.Samfile(bam_fname, "rb")
ref = pysam.Fastafile(ref_fname)
if not name:
name = os.path.basename(bam_fname)
if num_alleles:
print "# %s" % num_alleles
sys.stdout.write('\t'.join("chrom pos refbase altbase total refcount altcount background refback altback".split()))
if num_alleles and rscript:
sys.stdout.write("\tci_low\tci_high\tallele_lowt\tallele_high")
sys.stdout.write('\n')
for pileup in bam_pileup_iter(bam, mask=1540):
chrom = bam.getrname(pileup.tid)
counts = {'A': 0, 'C': 0, 'G': 0, 'T': 0}
total = 0
for pileupread in pileup.pileups:
if not pileupread.is_del:
if min_qual:
if pileupread.alignment.qual[pileupread.qpos] < min_qual:
continue
if pileupread.indel == 0:
base = pileupread.alignment.seq[pileupread.qpos].upper()
if base != 'N':
counts[base] += 1
total += 1
if total > min_count:
refbase = ref.fetch(chrom, pileup.pos, pileup.pos + 1).upper()
if not refbase in counts:
continue
refcount = counts[refbase]
# sort non-ref counts. first is alt, next is background
scounts = []
for c in counts:
if c != refbase:
scounts.append((counts[c], c))
scounts.sort()
scounts.reverse()
altbase = scounts[0][1]
altcount = scounts[0][0]
background = scounts[1][0]
refback = refcount - background
altback = altcount - background
if (altback + refback) == 0:
altfreq = 0
else:
altfreq = float(altback) / (altback + refback)
cols = [chrom, pileup.pos + 1, refbase, altbase, total, refcount, altcount, background, refback, altback, altfreq]
if num_alleles and rscript:
ci_low, ci_high = calc_cp_ci(refback + altback, altback, num_alleles)
allele_low = ci_low * num_alleles
allele_high = ci_high * num_alleles
cols.append(ci_low)
cols.append(ci_high)
cols.append(allele_low)
cols.append(allele_high)
else:
ci_low = 0
if not math.isnan(ci_low) and (min_ci_low is None or ci_low > min_ci_low):
print '\t'.join([str(x) for x in cols])
bam.close()
ref.close()
def bam_cims_finder(bam_fnames,
output='bed',
ref_fname=None,
flanking=12,
cutoff=0.1,
stranded=True,
window_size=20):
for bam_fname in bam_fnames:
sys.stderr.write('%s\n' % bam_fname)
bam = pysam.Samfile(bam_fname, "rb")
if output == 'fasta':
emitter = FASTAEmitter(ref_fname, flanking)
else:
emitter = BEDEmitter()
if stranded:
strands = ['+', '-']
else:
strands = ['']
for strand in strands:
manager = RegionManager(emitter, strand, window_size)
for pileup in bam_pileup_iter(bam, mask=1540):
chrom = bam.getrname(pileup.tid)
deletions = 0.0
total = 0.0
del_reads = set()
total_reads = set()
for pileupread in pileup.pileups:
if not strand or (strand == '+' and
not pileupread.alignment.is_reverse) or (
strand == '-'
and pileupread.alignment.is_reverse):
if is_read_match_at_pos(pileupread.alignment,
pileupread.qpos):
total += 1
total_reads.add(pileupread.alignment.qname)
if is_read_del_at_pos(pileupread.alignment,
pileupread.qpos):
deletions += 1
del_reads.add(pileupread.alignment.qname)
# print ""
# print chrom
# print pileup.pos
# print pileupread.alignment.qname
# print pileupread.alignment.pos
# print pileupread.alignment.cigar
# print pileupread.qpos
if total > 0:
pct = deletions / total
if pct > cutoff:
manager.add(chrom, pileup.pos, strand, del_reads,
total_reads)
manager.close()
bam.close()
emitter.close()
def bam_tobedgraph(bamfile, strand=None, normalize=None, nogaps=False, out=sys.stdout):
last_chrom = None
last_count = 0
last_start = None
last_end = None
for pileup in bam_pileup_iter(bamfile):
# sys.stdin.readline()
chrom = bamfile.getrname(pileup.tid)
if chrom != last_chrom and last_count > 0 and last_end:
write_bedgraph(last_chrom, last_start, last_end + 1, last_count, normalize, out)
last_count = 0
last_start = None
last_end = None
count = 0
if strand is None:
if not nogaps:
count = pileup.n
else:
for read in pileup.pileups:
op = read_cigar_at_pos(read.alignment.cigar, read.qpos, read.is_del)
if op != 3:
count += 1
else:
#
# TODO - add rev_read2 option
#
for read in pileup.pileups:
if not read.alignment.is_reverse and strand == '+':
if not nogaps:
count += 1
else:
op = read_cigar_at_pos(read.alignment.cigar, read.qpos, read.is_del)
if op != 3:
count += 1
elif read.alignment.is_reverse and strand == '-':
if not nogaps:
count += 1
else:
op = read_cigar_at_pos(read.alignment.cigar, read.qpos, read.is_del)
if op != 3:
count += 1
# print pileup.pos,count,last_start,last_end
if count != last_count or not last_end or (pileup.pos - last_end) > 1:
if last_count > 0:
write_bedgraph(last_chrom, last_start, last_end + 1, last_count, normalize, out)
if count == 0:
last_start = None
# elif not last_end or (pileup.pos-last_end) > 1:
# last_start = pileup.pos
else:
last_start = pileup.pos
last_end = pileup.pos
last_count = count
last_chrom = chrom
write_bedgraph(last_chrom, last_start, last_end + 1, last_count, normalize, out)
