def main():
parser = argparse.ArgumentParser(description="Make sam file compatible with tools counting on a splicemap format sam file.")
parser.add_argument('in_sam',help="FILENAME of sam file, or '-' for STDIN")
args = parser.parse_args()
inf = sys.stdin
if args.in_sam != '-':
inf = open(args.in_sam)
for line in inf:
line = line.rstrip()
if SamBasics.is_header(line):
print line
continue
f = line.rstrip().split("\t")
e = SamBasics.sam_line_to_dictionary(line)
if SamBasics.check_flag(e['flag'],4):
continue # skip the unmapped reads
if SamBasics.check_flag(e['flag'],16):
f[1] = "16"
else:
f[1] = "0"
f[4] = "0"
f[6] = "*"
f[7] = "0"
f[8] = "0"
print "\t".join(f)
def main():
parser = argparse.ArgumentParser(
description=
"Make sam file compatible with tools counting on a splicemap format sam file."
)
parser.add_argument('in_sam',
help="FILENAME of sam file, or '-' for STDIN")
args = parser.parse_args()
inf = sys.stdin
if args.in_sam != '-':
inf = open(args.in_sam)
for line in inf:
line = line.rstrip()
if SamBasics.is_header(line):
print line
continue
f = line.rstrip().split("\t")
e = SamBasics.sam_line_to_dictionary(line)
if SamBasics.check_flag(e['flag'], 4):
continue # skip the unmapped reads
if SamBasics.check_flag(e['flag'], 16):
f[1] = "16"
else:
f[1] = "0"
f[4] = "0"
f[6] = "*"
f[7] = "0"
f[8] = "0"
print "\t".join(f)
def get_exons_from_seqs(seqs, d, spcf):
sind = 0
oline = ''
for seq in seqs:
sind += 1
psec = 'P' #primary or secondary
if sind > 1: psec = 'S'
d1 = d.copy()
d1['rname'] = seq[1]
if seq[2] == '+': d1['flag'] = 0
else: d1['flag'] = 16
d1['pos'] = seq[3]
d1['cigar'] = seq[4]
d1['cigar_array'] = SamBasics.parse_cigar(seq[4])
skips = set(['H', 'D', 'N'])
total_length = 0
possible_matches = 0
indels = 0
qstart = 0
if d1['cigar_array'][0]['op'] == 'S':
qstart = d1['cigar_array'][0]['val']
if d1['cigar_array'][0]['op'] == 'H':
qstart = d1['cigar_array'][0]['val']
for ce in d1['cigar_array']:
if ce['op'] not in skips:
total_length += ce['val']
if ce['op'] == 'M': possible_matches += ce['val']
elif ce['op'] == 'I':
indels += ce['val']
elif ce['op'] == 'D' and ce['val'] < 68:
indels += ce['val']
fakeseq = 'N' * total_length
d1['seq'] = fakeseq
nline = SamBasics.entry_to_line(d1)
pline = spcf.convert_line(nline)
pentry = PSLBasics.line_to_entry(pline)
#mismatch_count = -1
#if sind == 1 and args.reference_genome: #for primary alignments we can calculate the number of matches
# for i in range(0,len(pentry['blockSizes'])):
# tseq = spcf.genome[pentry['tName']][pentry['tStarts'][i]:pentry['tStarts'][i]+pentry['blockSizes'][i]]
# qseq = sequence[pentry['qStarts'][i]:pentry['qStarts'][i]+pentry['blockSizes'][i]]
# print pentry['blockSizes'][i]
# print tseq
# print qseq
# for j in range(0,len(tseq)):
# if tseq[j].upper() != qseq[j].upper(): mismatch_count += 1
gline = PSLBasics.convert_entry_to_genepred_line(pentry)
gentry = GenePredBasics.line_to_entry(gline)
gsmooth = GenePredBasics.smooth_gaps(gentry, 68)
for i in range(0, len(gsmooth['exonStarts'])):
oline += gsmooth['chrom'] + "\t" + str(
gsmooth['exonStarts'][i]) + "\t" + str(
gsmooth['exonEnds']
[i]) + "\t" + gsmooth['strand'] + "\t" + gsmooth[
'name'] + "\t" + str(possible_matches) + "\t" + str(
indels) + "\t" + psec + "\t" + str(qstart) + "\n"
return oline
def get_exons_from_seqs(seqs,d,spcf):
sind = 0
oline = ''
for seq in seqs:
sind+=1
psec = 'P' #primary or secondary
if sind > 1: psec = 'S'
d1 = d.copy()
d1['rname'] = seq[1]
if seq[2] == '+': d1['flag'] = 0
else: d1['flag'] = 16
d1['pos'] = seq[3]
d1['cigar'] = seq[4]
d1['cigar_array'] = SamBasics.parse_cigar(seq[4])
skips = set(['H','D','N'])
total_length = 0
possible_matches = 0
indels = 0
qstart = 0
if d1['cigar_array'][0]['op'] == 'S':
qstart = d1['cigar_array'][0]['val']
if d1['cigar_array'][0]['op'] == 'H':
qstart = d1['cigar_array'][0]['val']
for ce in d1['cigar_array']:
if ce['op'] not in skips:
total_length += ce['val']
if ce['op'] == 'M': possible_matches += ce['val']
elif ce['op'] == 'I':
indels += ce['val']
elif ce['op'] == 'D' and ce['val'] < 68:
indels += ce['val']
fakeseq = 'N'*total_length
d1['seq'] = fakeseq
nline = SamBasics.entry_to_line(d1)
pline = spcf.convert_line(nline)
pentry = PSLBasics.line_to_entry(pline)
#mismatch_count = -1
#if sind == 1 and args.reference_genome: #for primary alignments we can calculate the number of matches
# for i in range(0,len(pentry['blockSizes'])):
# tseq = spcf.genome[pentry['tName']][pentry['tStarts'][i]:pentry['tStarts'][i]+pentry['blockSizes'][i]]
# qseq = sequence[pentry['qStarts'][i]:pentry['qStarts'][i]+pentry['blockSizes'][i]]
# print pentry['blockSizes'][i]
# print tseq
# print qseq
# for j in range(0,len(tseq)):
# if tseq[j].upper() != qseq[j].upper(): mismatch_count += 1
gline = PSLBasics.convert_entry_to_genepred_line(pentry)
gentry = GenePredBasics.line_to_entry(gline)
gsmooth = GenePredBasics.smooth_gaps(gentry,68)
for i in range(0,len(gsmooth['exonStarts'])):
oline += gsmooth['chrom'] + "\t" + str(gsmooth['exonStarts'][i])+"\t"+str(gsmooth['exonEnds'][i])+"\t"+gsmooth['strand']+"\t"+gsmooth['name']+"\t"+str(possible_matches)+"\t"+str(indels)+"\t"+psec+"\t"+str(qstart)+"\n"
return oline
def make_exons(args, thread_index, thread_count):
is_sam = True
if re.search('\.bam$', args.sam_file):
is_sam = False
stag = ''
if is_sam: stag = '-S'
cmd = 'samtools view -F 4 ' + stag + ' ' + args.sam_file
spcf = SamBasics.SAMtoPSLconversionFactory()
if args.reference_genome:
spcf.set_genome(args.reference_genome)
sampipe = subprocess.Popen(cmd, shell=True, stdout=subprocess.PIPE)
fname = args.tempdir + '/bedpart.' + str(thread_index) + '.bed'
of = open(fname, 'w')
z = 0
with sampipe.stdout as inf:
for line in inf:
z += 1
if z % thread_count != thread_index: continue
line = line.rstrip()
if SamBasics.is_header(line):
continue
d = SamBasics.sam_line_to_dictionary(line)
strand = '+'
if SamBasics.check_flag(d['flag'], 16):
strand = '-'
seqs = []
sequence = d['seq']
seqs.append([d['qname'], d['rname'], strand, d['pos'], d['cigar']])
m = re.search('XA:Z:(\S+)', line)
if m and args.use_secondary_alignments:
e = m.group(1)
secondaries = e.rstrip(";").split(";")
for secondary in secondaries:
m1 = re.match('([^,]+),([+-])(\d+),([^,]+)', secondary)
if not m1:
sys.stderr.write("strange secondary format " +
secondary + "\n")
sys.exit()
seqs.append([
d['qname'],
m1.group(1),
m1.group(2),
int(m1.group(3)),
m1.group(4)
])
#p.apply_async(get_exons_from_seqs,[seqs,d,spcf])
exons = get_exons_from_seqs(seqs, d, spcf)
of.write(exons)
#return exons
of.close()
def main():
parser = argparse.ArgumentParser(
description="Get read counts from sam or bam.")
parser.add_argument('input', help="FILENAME sam or bam")
parser.add_argument(
'--add_report',
action='store_true',
help="make a new file where we replace sam or bam with a .mapped_count"
)
args = parser.parse_args()
if args.add_report:
m = re.match('(.+)\.[bs]am', args.input)
if not m:
sys.stderr.write("bad inputfile type should be .bam or .sam\n")
sys.exit()
baseinput = m.group(1)
samtag = ''
if re.search('\.sam$', args.input): samtag = '-S'
z = 0
#se = open('/dev/stderr','w')
p = subprocess.Popen('sort | uniq | wc -l',
shell=True,
stdin=subprocess.PIPE,
stdout=subprocess.PIPE)
with os.popen('samtools view ' + samtag + ' ' + args.input) as inf:
for line in inf:
z += 1
if z % 100000 == 0:
sys.stderr.write(str(z) + " alignments processed\r")
line = line.rstrip()
d = SamBasics.sam_line_to_dictionary(line)
if not SamBasics.check_flag(d['flag'], 4):
if SamBasics.check_flag(d['flag'], 64):
p.stdin.write(d['qname'] + '.1' + "\n")
elif SamBasics.check_flag(d['flag'], 128):
p.stdin.write(d['qname'] + '.2' + "\n")
else:
sys.stderr.write("Unrecognized\n")
sys.exit()
sys.stderr.write("\n")
aligned_reads = int(p.communicate()[0].rstrip())
if args.add_report:
of = open(baseinput + '.mapped_reads', 'w')
of.write(str(aligned_reads) + "\n")
return
print aligned_reads
def main():
parser = argparse.ArgumentParser(
description=
"Find mapping distance of paired end reads. Takes an ordered (by query) alignment to a transcriptome.\nSomething that works for an input thus far is like:\nhisat --reorder -x mytranscriptome -1 my_1.fastq -2 my_2.fastq | this_script.py -"
)
parser.add_argument(
'input_sam',
help="SAMFILE ordered alignment a transcriptome or - for stdin")
args = parser.parse_args()
inf = sys.stdin
if args.input_sam != '-':
inf = open(args.input_sam)
msr = SamBasics.MultiEntrySamReader(inf)
spcf = SamBasics.SAMtoPSLconversionFactory()
data = []
sys.stderr.write("Pairs Mean Stddev\n")
while True:
entries = msr.read_entries()
if not entries: break
if len(entries) != 2: continue
[e1, e2] = entries
if e1.check_flag(4) or e2.check_flag(4): continue
if not e1.check_flag(2) and e2.check_flag(2): continue
if not ((e1.check_flag(64) and e2.check_flag(128)) or
(e1.check_flag(128) and e2.check_flag(64))):
continue
p1 = spcf.convert_line(e1.get_line())
p2 = spcf.convert_line(e2.get_line())
if not p1 or not p2: continue
p1 = PSLBasics.PSL(p1)
p2 = PSLBasics.PSL(p2)
dist = max(
p2.value('tEnd') - p1.value('tStart'),
p1.value('tEnd') - p2.value('tStart'))
data.append(dist)
if len(data) < 2: continue
if len(data) % 1000 == 0:
sys.stderr.write(
str(len(data)) + " " + str(int(mean(data))) + " " +
str(int(stddev(data))) + " \r")
sys.stderr.write(
str(len(data)) + " " + str(int(mean(data))) + " " +
str(int(stddev(data))) + " \r")
sys.stderr.write("\n")
def make_exons(args,thread_index,thread_count):
is_sam = True
if re.search('\.bam$',args.sam_file):
is_sam = False
stag = ''
if is_sam: stag = '-S'
cmd = 'samtools view -F 4 '+stag+' '+args.sam_file
spcf = SamBasics.SAMtoPSLconversionFactory()
if args.reference_genome:
spcf.set_genome(args.reference_genome)
sampipe = subprocess.Popen(cmd,shell=True,stdout=subprocess.PIPE)
fname = args.tempdir+'/bedpart.'+str(thread_index)+'.bed'
of = open(fname,'w')
z = 0
with sampipe.stdout as inf:
for line in inf:
z+=1
if z%thread_count != thread_index: continue
line = line.rstrip()
if SamBasics.is_header(line):
continue
d = SamBasics.sam_line_to_dictionary(line)
strand = '+'
if SamBasics.check_flag(d['flag'],16):
strand = '-'
seqs = []
sequence = d['seq']
seqs.append([d['qname'], d['rname'], strand, d['pos'], d['cigar']])
m = re.search('XA:Z:(\S+)',line)
if m and args.use_secondary_alignments:
e = m.group(1)
secondaries = e.rstrip(";").split(";")
for secondary in secondaries:
m1 = re.match('([^,]+),([+-])(\d+),([^,]+)',secondary)
if not m1:
sys.stderr.write("strange secondary format "+secondary+"\n")
sys.exit()
seqs.append([d['qname'], m1.group(1),m1.group(2),int(m1.group(3)),m1.group(4)])
#p.apply_async(get_exons_from_seqs,[seqs,d,spcf])
exons = get_exons_from_seqs(seqs,d,spcf)
of.write(exons)
#return exons
of.close()
def main():
parser = argparse.ArgumentParser()
parser.add_argument('psl',help="FILENAME of psl file (can be gzipped)")
parser.add_argument('refgenome',help="FASTA of the reference genome")
parser.add_argument('--min_intron_size',default=68,type=int,help="INT minimum intron size")
parser.add_argument('--fastq_reads',help="FASTQ of the reads")
parser.add_argument('--fasta_reads',help="FASTA of the reads")
parser.add_argument('--skip_directionless_splice',action='store_true',help='only output reads where canonical splice sites indicate direciton if junctions are present')
parser.add_argument('-o',help="FILENAME to save sam output")
args = parser.parse_args()
pscf = SamBasics.PSLtoSAMconversionFactory()
pscf.set_min_intron_size(args.min_intron_size)
sys.stderr.write("Creating header from reference fasta\n")
#header = SamBasics.construct_header_from_reference_fasta('/Shared/Au/jason/Reference/UCSC/Human/hg19_GRCh37_feb2009/Genome/genome.fa')
if args.skip_directionless_splice:
pscf.set_skip_directionless_splice()
header = SamBasics.construct_header_from_reference_fasta(args.refgenome)
if args.o:
of = open(args.o,'w')
of.write(header)
else:
sys.stdout.write(header)
sys.stderr.write("setting reference fasta for conversion\n")
pscf.set_reference_genome(args.refgenome)
sys.stderr.write("determining mapping counts from psl\n")
pscf.set_mapping_counts(args.psl)
#pscf.construct_header_from_reference_fasta('test.fa')
sys.stderr.write("Establishing library of reads\n")
if args.fastq_reads:
pscf.set_read_fastq(args.fastq_reads)
elif args.fasta_reads:
pscf.set_read_fasta(args.fasta_reads)
sys.stderr.write("Performing conversion\n")
gfr = None
if args.psl[-3:]=='.gz':
gfr = gzip.open(args.psl)
else:
gfr = open(args.psl)
skipped = 0
while True:
line = gfr.readline()
if not line: break
samline = pscf.convert_line(line.rstrip())
if not samline:
skipped += 1
sys.stderr.write("\rskipping directionless splice ("+str(skipped)+") ")
continue # happens if we are skipping directionless splice
if args.o:
of.write(samline+"\n")
else:
sys.stdout.write(samline+"\n")
if args.o:
of.close()
gfr.close()
sys.stderr.write("\n")
def main():
parser = argparse.ArgumentParser(description="Get read counts from sam or bam.")
parser.add_argument('input',help="FILENAME sam or bam")
parser.add_argument('--add_report',action='store_true',help="make a new file where we replace sam or bam with a .mapped_count")
args = parser.parse_args()
if args.add_report:
m = re.match('(.+)\.[bs]am',args.input)
if not m:
sys.stderr.write("bad inputfile type should be .bam or .sam\n")
sys.exit()
baseinput = m.group(1)
samtag = ''
if re.search('\.sam$',args.input): samtag = '-S'
z = 0
#se = open('/dev/stderr','w')
p = subprocess.Popen('sort | uniq | wc -l',shell=True,stdin=subprocess.PIPE,stdout=subprocess.PIPE)
with os.popen('samtools view '+samtag+' '+args.input) as inf:
for line in inf:
z += 1
if z%100000 ==0:
sys.stderr.write(str(z)+" alignments processed\r")
line = line.rstrip()
d = SamBasics.sam_line_to_dictionary(line)
if not SamBasics.check_flag(d['flag'],4):
if SamBasics.check_flag(d['flag'],64):
p.stdin.write(d['qname']+'.1'+"\n")
elif SamBasics.check_flag(d['flag'],128):
p.stdin.write(d['qname']+'.2'+"\n")
else:
sys.stderr.write("Unrecognized\n")
sys.exit()
sys.stderr.write("\n")
aligned_reads = int(p.communicate()[0].rstrip())
if args.add_report:
of = open(baseinput+'.mapped_reads','w')
of.write(str(aligned_reads)+"\n")
return
print aligned_reads
def read_sam_file(self,filename):
gsr = SamBasics.GenericSamReader(filename)
linecount = 0
while True and linecount < self.max_read_count:
line1 = gsr.readline().rstrip()
if not line1: break
line2 = gsr.readline().rstrip()
if not line2: break
line3 = gsr.readline().rstrip()
if not line3: break
line4 = gsr.readline().rstrip()
if not line4: break
self.record_observation(line4)
linecount += 1
gsr.close()
def main():
parser = argparse.ArgumentParser(description="Read a sam file and output a bed file in the format of junction_color.bed")
parser.add_argument('-o','--output',help='FILENAME is output')
parser.add_argument('--min_intron_size',type=int,default=68,help='minimum intron size')
parser.add_argument('infile',help='FILENAME of sam file or "-" for STDIN')
parser.add_argument('reference_genome',help='FILENAME of the reference genome')
args = parser.parse_args()
# get our reference genome
sys.stderr.write("reading reference genome\n")
#g = SequenceBasics.read_fasta_into_hash(args.reference_genome)
g = FastaData(open(args.reference_genome).read())
sys.stderr.write("finished reading reference genome\n")
inf = sys.stdin
read_mapping_count = {}
junctions = {}
if args.infile != '-':
inf = open(args.infile)
sys.stderr.write("reading through sam file\n")
zall = 0
zn = 0
while True:
line = inf.readline()
if not line: break
line = line.rstrip()
if SamBasics.is_header(line): continue
d = SamBasics.sam_line_to_dictionary(line)
chrom = d['rname']
if chrom =='*': continue
if chrom not in g.keys():
sys.stderr.write("WARNING: "+chrom+" not in reference, skipping\n")
continue
mate = 'U'
if SamBasics.check_flag(d['flag'],int('0x4',16)): #check if its unmapped
continue # we can ignore the unmapped things for now
if SamBasics.check_flag(d['flag'],int('0x40',16)):
mate = 'L'
elif SamBasics.check_flag(d['flag'],int('0x80',16)):
mate = 'R'
actual_read = d['qname']+"\t"+mate
if actual_read not in read_mapping_count:
read_mapping_count[actual_read] = 0
read_mapping_count[actual_read] += 1
has_intron = 0
start_loc = d['pos']
current_loc = start_loc
bounds = []
for i in range(0,len(d['cigar_array'])):
ce = d['cigar_array'][i]
if ce['op'] == 'N' and ce['val'] >= args.min_intron_size:
has_intron = 1
lbound = current_loc # should be the intron start base index-1
current_loc += ce['val']
rbound = current_loc # should be the second exon start base index-1
right_size = d['cigar_array'][i+1]['val']
bounds.append([lbound,rbound,right_size])
elif ce['op'] == 'D':
current_loc += ce['val']
elif re.match('[=XMSHP]',ce['op']):
current_loc += ce['val']
if has_intron == 0: continue # there are no splices to report here
#print actual_read
#print d['cigar']
#print d
#print start_loc
#print bounds
for bound in bounds:
zall += 1
intronflank = g[chrom][bound[0]-1:bound[0]+1].upper() + '-' + \
g[chrom][bound[1]-3:bound[1]-1].upper()
strand = ''
if is_canon(intronflank): # its a positive strand
strand = '+'
elif is_revcanon(intronflank): # its a negative strand
strand = '-'
else:
# We can't deal with the non-canonical splice sorry
zn += 1
sys.stderr.write("WARNING skipping non-canonical splice ("+str(zn)+"/"+str(zall)+")\r")
continue
# If we are still in we have successfully found a splice
out_chrom = chrom
out_start = bound[0]-51
out_end = bound[1]+49
out_name = '*' # this will be done later
out_score = 50
out_strand = strand
out_thickStart = out_start
out_thickEnd = out_end
out_rgb = '0,0,0'
out_block_count = 2
out_block_sizes = '50,50'
out_block_starts = '0,'+str(bound[1]-bound[0]+50)
bed = []
bed.append(out_chrom)
bed.append(str(out_start))
bed.append(str(out_end))
bed.append(out_name)
bed.append(str(out_score))
bed.append(out_strand)
bed.append(str(out_thickStart))
bed.append(str(out_thickEnd))
bed.append(out_rgb)
bed.append(str(out_block_count))
bed.append(out_block_sizes)
bed.append(out_block_starts)
entry = "\t".join(bed)
if entry not in junctions:
junctions[entry] = {}
junctions[entry]['reads'] = set()
junctions[entry]['positions'] = set()
junctions[entry]['right_sizes'] = set()
junctions[entry]['reads'].add(actual_read)
junctions[entry]['positions'].add(d['pos'])
junctions[entry]['right_sizes'].add(bound[2])
sys.stderr.write("\n")
sys.stderr.write("finished reading sam\n")
of = sys.stdout
if args.output:
of = open(args.output,'w')
if len(junctions) > 0: # if we have stuff lets print a header
of.write("track\tname=junctions\tdescription=\"SpliceMap junctions\" itemRgb=\"On\"\n")
for entry in junctions:
nR = len(junctions[entry]['reads'])
width = max(junctions[entry]['right_sizes'])-min(junctions[entry]['right_sizes'])
nNR = len(junctions[entry]['positions'])
nUR = 0
nMR = 0
for read in junctions[entry]['reads']:
if read_mapping_count[read] == 1:
nUR += 1
elif read_mapping_count[read] > 1:
nMR += 1
else:
sys.stderr.write("ERROR: nonsense read count\n")
return
name = '('+str(nR)+')['+str(width)+'_'+str(nNR)+']('+str(nUR)+'/'+str(nMR)+')'
bed = entry.split("\t")
bed[3] = name
of.write("\t".join(bed)+"\n")
def main():
parser = argparse.ArgumentParser(
description="Convert a sam file into a psl file")
parser.add_argument('--genome',
help="FASTA input file of reference genome")
parser.add_argument('--get_secondary_alignments',
action='store_true',
help="Report SA:Z secondary alignments as well")
parser.add_argument('--get_alternative_alignments',
action='store_true',
help="Report XA:Z alternative alignments as well")
parser.add_argument(
'--get_all_alignments',
action='store_true',
help="Report SA:Z and XA:Z alternative alignments as well")
parser.add_argument('--give_unique_names',
action='store_true',
help="Output query names will be unique.")
group = parser.add_mutually_exclusive_group()
group.add_argument(
'--output_fasta',
help=
"FILENAME to save an outgoing fasta. Only works for primary alignments."
)
group.add_argument(
'--output_fastq',
help=
"FILENAME to save an outgoing fastq. Only works for primary alignments."
)
parser.add_argument('infile', help="FILENAME input file or '-' for STDIN")
parser.add_argument('-o',
'--output',
help="FILENAME for the output, STDOUT if not set.")
args = parser.parse_args()
if (args.output_fasta
or args.output_fastq) and (args.get_secondary_alignments
or args.get_alternative_alignments
or args.get_all_alignments):
sys.stderr.write(
"ERROR, can only output the fastq/fasta if we are doing primary alignments only.\n"
)
sys.exit()
inf = sys.stdin
if args.infile != '-':
inf = open(args.infile)
of = sys.stdout
if args.output:
of = open(args.output, 'w')
spcf = SamBasics.SAMtoPSLconversionFactory()
if args.genome: spcf.set_genome(args.genome)
off = None
if args.output_fasta:
off = open(args.output_fasta, 'w')
if args.output_fastq:
off = open(args.output_fastq, 'w')
z = 0
for line in inf:
line = line.rstrip()
if SamBasics.is_header(line):
spcf.read_header_line(line)
continue
# We have a line to convert
psl = spcf.convert_line(line)
if psl:
pobj = PSL(psl)
z += 1
if args.give_unique_names:
pobj.entry['qName'] = 'Q' + str(z)
of.write(pobj.get_line() + "\n")
if args.output_fastq or args.output_fasta:
sam = SamBasics.SAM(line)
sequence = sam.value('seq').upper()
quality = sam.value('qual')
if sam.check_flag(16):
sequence = rc(sam.value('seq').upper())
quality = sam.value('qual')[::-1]
if args.output_fasta:
off.write(">" + pobj.value('qName') + "\n" + sequence +
"\n")
elif args.output_fastq:
if len(sequence) == len(quality):
off.write("@" + pobj.value('qName') + "\n" + sequence +
"\n" + "+\n" + quality + "\n")
else:
sys.stderr.write("ERROR: sequence " + sequence +
" length (" + str(len(sequence)) +
") doesnt match quality " + quality +
" length (" + str(len(quality)) +
")\n")
sys.exit()
# Lets look for secondary alignments to convert
if args.get_secondary_alignments or args.get_all_alignments:
secondary_alignments = SamBasics.get_secondary_alignments(
line.rstrip())
for samline in secondary_alignments:
psl = spcf.convert_line(samline)
if psl:
#print "\nsecondary"
#print samline
z += 1
pobj = PSL(psl)
if args.give_unique_names:
pobj.entry['qName'] = 'Q' + str(z)
of.write(pobj.get_line() + "\n")
if args.get_alternative_alignments or args.get_all_alignments:
alternative_alignments = SamBasics.get_alternative_alignments(
line.rstrip())
for samline in alternative_alignments:
psl = spcf.convert_line(samline)
if psl:
#print "\nsecondary"
#print samline
z += 1
pobj = PSL(psl)
if args.give_unique_names:
pobj.entry['qName'] = 'Q' + str(z)
of.write(pobj.get_line() + "\n")
inf.close()
of.close()
def main():
parser = argparse.ArgumentParser(description="Convert a sam file into a psl file")
parser.add_argument('--genome',help="FASTA input file of reference genome")
parser.add_argument('--get_secondary_alignments',action='store_true',help="Report SA:Z secondary alignments as well")
parser.add_argument('--get_alternative_alignments',action='store_true',help="Report XA:Z alternative alignments as well")
parser.add_argument('--get_all_alignments',action='store_true',help="Report SA:Z and XA:Z alternative alignments as well")
parser.add_argument('--give_unique_names',action='store_true',help="Output query names will be unique.")
group = parser.add_mutually_exclusive_group()
group.add_argument('--output_fasta',help="FILENAME to save an outgoing fasta. Only works for primary alignments.")
group.add_argument('--output_fastq',help="FILENAME to save an outgoing fastq. Only works for primary alignments.")
parser.add_argument('infile',help="FILENAME input file or '-' for STDIN")
parser.add_argument('-o','--output',help="FILENAME for the output, STDOUT if not set.")
args = parser.parse_args()
if (args.output_fasta or args.output_fastq) and (args.get_secondary_alignments or args.get_alternative_alignments or args.get_all_alignments):
sys.stderr.write("ERROR, can only output the fastq/fasta if we are doing primary alignments only.\n")
sys.exit()
inf = sys.stdin
if args.infile != '-':
inf = open(args.infile)
of = sys.stdout
if args.output:
of = open(args.output,'w')
spcf = SamBasics.SAMtoPSLconversionFactory()
if args.genome: spcf.set_genome(args.genome)
off = None
if args.output_fasta:
off = open(args.output_fasta,'w')
if args.output_fastq:
off = open(args.output_fastq,'w')
z = 0
for line in inf:
line = line.rstrip()
if SamBasics.is_header(line):
spcf.read_header_line(line)
continue
# We have a line to convert
psl = spcf.convert_line(line)
if psl:
pobj = PSL(psl)
z += 1
if args.give_unique_names:
pobj.entry['qName'] = 'Q'+str(z)
of.write(pobj.get_line()+"\n")
if args.output_fastq or args.output_fasta:
sam = SamBasics.SAM(line)
sequence = sam.value('seq').upper()
quality = sam.value('qual')
if sam.check_flag(16):
sequence = rc(sam.value('seq').upper())
quality = sam.value('qual')[::-1]
if args.output_fasta:
off.write(">"+pobj.value('qName')+"\n"+sequence+"\n")
elif args.output_fastq:
if len(sequence) == len(quality):
off.write("@"+pobj.value('qName')+"\n"+sequence+"\n"+"+\n"+quality+"\n")
else:
sys.stderr.write("ERROR: sequence "+sequence+" length ("+str(len(sequence))+") doesnt match quality "+quality+" length ("+str(len(quality))+")\n")
sys.exit()
# Lets look for secondary alignments to convert
if args.get_secondary_alignments or args.get_all_alignments:
secondary_alignments = SamBasics.get_secondary_alignments(line.rstrip())
for samline in secondary_alignments:
psl = spcf.convert_line(samline)
if psl:
#print "\nsecondary"
#print samline
z += 1
pobj = PSL(psl)
if args.give_unique_names:
pobj.entry['qName'] = 'Q'+str(z)
of.write(pobj.get_line()+"\n")
if args.get_alternative_alignments or args.get_all_alignments:
alternative_alignments = SamBasics.get_alternative_alignments(line.rstrip())
for samline in alternative_alignments:
psl = spcf.convert_line(samline)
if psl:
#print "\nsecondary"
#print samline
z += 1
pobj = PSL(psl)
if args.give_unique_names:
pobj.entry['qName'] = 'Q'+str(z)
of.write(pobj.get_line()+"\n")
inf.close()
of.close()
def main():
parser = argparse.ArgumentParser()
parser.add_argument('psl', help="FILENAME of psl file (can be gzipped)")
parser.add_argument('refgenome', help="FASTA of the reference genome")
parser.add_argument('--min_intron_size',
default=68,
type=int,
help="INT minimum intron size")
parser.add_argument('--fastq_reads', help="FASTQ of the reads")
parser.add_argument('--fasta_reads', help="FASTA of the reads")
parser.add_argument(
'--skip_directionless_splice',
action='store_true',
help=
'only output reads where canonical splice sites indicate direciton if junctions are present'
)
parser.add_argument('-o', help="FILENAME to save sam output")
args = parser.parse_args()
pscf = SamBasics.PSLtoSAMconversionFactory()
pscf.set_min_intron_size(args.min_intron_size)
sys.stderr.write("Creating header from reference fasta\n")
#header = SamBasics.construct_header_from_reference_fasta('/Shared/Au/jason/Reference/UCSC/Human/hg19_GRCh37_feb2009/Genome/genome.fa')
if args.skip_directionless_splice:
pscf.set_skip_directionless_splice()
header = SamBasics.construct_header_from_reference_fasta(args.refgenome)
if args.o:
of = open(args.o, 'w')
of.write(header)
else:
sys.stdout.write(header)
sys.stderr.write("setting reference fasta for conversion\n")
pscf.set_reference_genome(args.refgenome)
sys.stderr.write("determining mapping counts from psl\n")
pscf.set_mapping_counts(args.psl)
#pscf.construct_header_from_reference_fasta('test.fa')
sys.stderr.write("Establishing library of reads\n")
if args.fastq_reads:
pscf.set_read_fastq(args.fastq_reads)
elif args.fasta_reads:
pscf.set_read_fasta(args.fasta_reads)
sys.stderr.write("Performing conversion\n")
gfr = None
if args.psl[-3:] == '.gz':
gfr = gzip.open(args.psl)
else:
gfr = open(args.psl)
skipped = 0
while True:
line = gfr.readline()
if not line: break
samline = pscf.convert_line(line.rstrip())
if not samline:
skipped += 1
sys.stderr.write("\rskipping directionless splice (" +
str(skipped) + ") ")
continue # happens if we are skipping directionless splice
if args.o:
of.write(samline + "\n")
else:
sys.stdout.write(samline + "\n")
if args.o:
of.close()
gfr.close()
sys.stderr.write("\n")
def check_parameters(z,gz,ifile,tdir,max_allowed_mismatches,Q,fsize):
#sys.stderr.write("doing "+str(z)+"\n")
g = {}
for n in gz:
g[n] = zlib.decompress(gz[n])
FNULL = open(os.devnull,'w')
cmd1 = "bwa mem "+ifile+" "+tdir+'/'+str(z)+'.fq'
cmd2 = "samtools view -S -"
stream1 = subprocess.Popen(cmd1.split(),stdout=subprocess.PIPE,stderr=FNULL)
stream2 = subprocess.Popen(cmd2.split(),stdin=stream1.stdout,stdout=subprocess.PIPE,stderr=FNULL)
reads = {}
while True:
sumlen= 0
mismatches = 0
line = stream2.stdout.readline()
if not line: break
f = line.rstrip().split("\t")
if f[2] == '*':
continue
d = SamBasics.sam_line_to_dictionary(line)
#if d['rname'] != 'chr20': continue #get rid of this line soon.
cigar = d['cigar_array']
#endmismatch = 0
#if cigar[0]['op'] == 'S':
# endmismatch += cigar[0]['val']
#if cigar[len(cigar)-1]['op'] == 'S':
# endmismatch += cigar[len(cigar)-1]['val']
#if endmismatch > max_end_mismatches: continue
read_index = 1
chrom_index = d['pos']
for e in cigar:
if re.match('[MX=]',e['op']):
sumlen += e['val'] # keep track of our match length
refseq = g[d['rname']][chrom_index-1:chrom_index-1+e['val']].upper()
readseq = d['seq'][read_index-1:read_index-1+e['val']].upper()
for i in range(0,e['val']):
if refseq[i] != readseq[i]: mismatches += 1
read_index += e['val']
chrom_index += e['val']
elif re.match('[SI]',e['op']):
mismatches += e['val']
read_index += e['val']
elif re.match('[NDH]',e['op']):
chrom_index += e['val']
else:
sys.stderr.write("warning: strange SAM op\n")
# save the biggest sum for the read name
#print 'mismatches: '+str(mismatches)
if mismatches > max_allowed_mismatches: continue
if d['qname'] not in reads:
reads[d['qname']] = {}
reads[d['qname']]['alignment_length'] = 0
reads[d['qname']]['mismatches'] = 0
if sumlen > reads[d['qname']]['alignment_length']:
reads[d['qname']]['alignment_length'] = sumlen
reads[d['qname']]['mismatches'] = mismatches
mapped_bases = 0
mapped_reads = 0
for rname in reads:
mapped_bases += reads[rname]['alignment_length']
mapped_reads += 1
#print str(mapped_bases) + "\t" + str(mapped_reads)
res = [z,mapped_reads,mapped_bases]
#sys.stderr.write(str(z)+"\t"+str(mapped_reads)+"\t"+str(mapped_bases)+"\n")
Q.put(res)
progress = Q.qsize()
sys.stderr.write('\r'+(' '*40))
sys.stderr.write('\r'+str(progress)+"/"+str(fsize))
sys.stderr.flush()
return