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 main():
parser = argparse.ArgumentParser(
description="Correct the matches/mismatches and Ncount of a PSL file")
parser.add_argument('input', help="PSLFILE or - for STIDN")
parser.add_argument('reference', help="FASTAFILE reference genome")
parser.add_argument('query', help="FASTAFILE query sequences")
parser.add_argument('--minimum_intron_size',
type=int,
default=68,
help="INT")
#parser.add_argument('--ordered_query',action='store_true',help="The query psl and fasta are both ordered by query name for optimal performance")
args = parser.parse_args()
# Read in the reference genome
sys.stderr.write("Reading in reference genome\n")
g = read_fasta_into_hash(args.reference)
sys.stderr.write("Finished reading " + str(len(g.keys())) +
" reference sequences\n")
inf = sys.stdin
if args.input != '-':
inf = open(args.input)
fhr = FastaHandleReader(open(args.query))
last_fasta = fhr.read_entry()
if not last_fasta:
sys.stderr.write("ERROR: No query sequences\n")
sys.exit()
for line in inf:
p = PSLBasics.PSL(line)
if not p.validate():
sys.stderr.write(
"WARNING skipping invalid PSL entry. This script fixes improper mismatch and match counts .. and gap counts... it doesn't perform miracles. Problem line:\n"
+ line.rstrip() + "\n")
n = p.value('qName')
if not last_fasta:
sys.stderr.write(
"ERROR: Ran out of query sequences too soon. Are they sorted properly\n"
)
sys.exit()
while last_fasta['name'] != n:
last_fasta = fhr.read_entry()
p.set_query(last_fasta['seq'])
p.set_reference_dictionary(g)
print p.get_line()
p.pretty_print(50)
fhr.close()
def main():
parser = argparse.ArgumentParser(description="Correct the matches/mismatches and Ncount of a PSL file")
parser.add_argument('input',help="PSLFILE or - for STIDN")
parser.add_argument('reference',help="FASTAFILE reference genome")
parser.add_argument('query',help="FASTAFILE query sequences")
parser.add_argument('--minimum_intron_size',type=int,default=68,help="INT")
#parser.add_argument('--ordered_query',action='store_true',help="The query psl and fasta are both ordered by query name for optimal performance")
args = parser.parse_args()
# Read in the reference genome
sys.stderr.write("Reading in reference genome\n")
g = read_fasta_into_hash(args.reference)
sys.stderr.write("Finished reading "+str(len(g.keys()))+" reference sequences\n")
inf = sys.stdin
if args.input != '-':
inf = open(args.input)
fhr = FastaHandleReader(open(args.query))
last_fasta = fhr.read_entry()
if not last_fasta:
sys.stderr.write("ERROR: No query sequences\n")
sys.exit()
for line in inf:
p = PSLBasics.PSL(line)
if not p.validate():
sys.stderr.write("WARNING skipping invalid PSL entry. This script fixes improper mismatch and match counts .. and gap counts... it doesn't perform miracles. Problem line:\n"+line.rstrip()+"\n")
n = p.value('qName')
if not last_fasta:
sys.stderr.write("ERROR: Ran out of query sequences too soon. Are they sorted properly\n")
sys.exit()
while last_fasta['name'] != n:
last_fasta = fhr.read_entry()
p.set_query(last_fasta['seq'])
p.set_reference_dictionary(g)
p.correct_stats()
print p.get_line()
continue
f = last_fasta
nCount = 0
matches = 0
misMatches = 0
prev_qE = 0
prev_tE = 0
qNumInsert = 0
qBaseInsert = 0
tNumInsert = 0
tBaseInsert = 0
for i in range(p.value('blockCount')):
blen = p.value('blockSizes')[i]
qS = p.value('qStarts')[i] #query start
qE = qS + blen #query end
tS = p.value('tStarts')[i] #target start
tE = tS + blen #target end
#Work on gaps
if prev_qE > 0 or prev_tE > 0: #if its not our first time through
tgap = tS-prev_tE
if tgap < args.minimum_intron_size and tgap > 0:
tNumInsert += 1
tBaseInsert += tgap
qgap = qS-prev_qE
if qgap > 0:
qNumInsert += 1
qBaseInsert += qgap
query = f['seq']
if p.value('strand') == '-':
query = rc(f['seq'])
qseq = query[qS:qE].upper()
rseq = g[p.value('tName')][tS:tE].upper()
#print qseq+"\n"+rseq+"\n"
for j in range(0,blen):
if qseq[j] == 'N':
nCount += 1
elif qseq[j] == rseq[j]:
matches += 1
else:
misMatches += 1
prev_qE = qE
prev_tE = tE
p.entry['matches'] = matches
p.entry['misMatches'] = misMatches
p.entry['nCount'] = nCount
p.entry['qNumInsert'] = qNumInsert
p.entry['qBaseInsert'] = qBaseInsert
p.entry['tNumInsert'] = tNumInsert
p.entry['tBaseInsert'] = tBaseInsert
p.entry['qSize'] = len(query)
p.entry['tSize'] = len(g[p.value('tName')])
print p.get_line()
#p.pretty_print(100)
fhr.close()
def main():
parser = argparse.ArgumentParser(
description="Analyze ORDERED psl alignments of long reads.")
parser.add_argument(
'psl_file',
help="Alignment file. Must be ordered by query name. use - for stdin")
parser.add_argument('-o',
'--output',
help="Write to output file, default is STDIN")
parser.add_argument('--noheader', action='store_true')
parser.add_argument(
'--minimum_coverage',
type=int,
help="Only consider alignments with at least this many bp aligned")
parser.add_argument('--threads',
type=int,
default=multiprocessing.cpu_count(),
help="INT default cpu_count")
parser.add_argument(
'--tempbuffer',
help=
"DIRECTORY store the results in a temporary file until they are ready to output. suggest using /tmp if you don't know what to use"
)
args = parser.parse_args()
seen_names = set()
last_name = ''
buffer = PSLBasics.MultiplePSLAlignments()
inf = sys.stdin
if args.psl_file != '-':
inf = open(args.psl_file)
global of
tname = None
if args.tempbuffer:
if not args.output:
sys.stderr.write(
"ERROR if you want to buffer outputs in a temp file you need to specify a final output file.\n"
)
sys.exit()
rnum = random.randint(1, 1000000000)
tname = args.tempbuffer.rstrip('/') + '/weirathe.' + str(
rnum) + '.meta'
of = open(tname, 'w')
if args.output and not args.tempbuffer:
of = open(args.output, 'w')
global lock
if args.threads > 1:
pool = multiprocessing.Pool(args.threads)
for line in inf:
e = PSLBasics.line_to_entry(line.rstrip())
if e['qName'] != last_name: # we have a new name
if e['qName'] in seen_names:
sys.stderr.write(
"ERROR psl entries are not ordered by query name.\n")
sys.exit()
seen_names.add(e['qName'])
if buffer.get_alignment_count() > 0:
#process_buffer(buffer)
if args.threads > 1:
pool.apply_async(process_buffer, [buffer],
callback=print_result)
else:
res = process_buffer(buffer)
print_result(res)
buffer = PSLBasics.MultiplePSLAlignments()
if args.minimum_coverage > 1:
buffer.set_minimum_coverage(args.minimum_coverage)
last_name = e['qName']
buffer.add_entry(PSLBasics.PSL(line.rstrip()))
inf.close()
if buffer.get_alignment_count() > 0:
if args.threads > 1:
pool.apply_async(
process_buffer, [buffer],
callback=print_result) # if we still have something left to do
else:
res = process_buffer(buffer)
print_result(res)
if args.threads > 1:
pool.close()
pool.join()
of.close()
if args.tempbuffer:
of = open(args.output, 'w')
with open(tname) as inf:
for line in inf:
of.write(line)
of.close()
os.remove(tname)