def set_read1_consensus_to_read2(input_stream, output_stream):
#get the header
line = input_stream.readline()
while line.startswith("@"):
output_stream.write(line)
line = input_stream.readline()
prev_read=Sam_record(line)
for line in input_stream:
read=Sam_record(line)
if prev_read and read.get_query_name() == prev_read.get_query_name():
if read.is_second_read() and prev_read.is_first_read():
read1=prev_read
read2=read
else:
read2=prev_read
read1=read
if not read1.is_unmapped():
read2.set_reference_name(read1.get_reference_name())
read2.set_unmapped_flag(False)
read2.set_position(1)
read2.set_cigar_string("%sM"%len(read2.get_query_sequence()))
output_stream.write(str(read1))
output_stream.write(str(read2))
prev_read=None
elif prev_read:
output_stream.write(str(prev_read))
prev_read=read
else:
prev_read=read
def set_read1_consensus_to_read1_and_read2(input_stream, output_stream):
#get the header
line = input_stream.readline()
while line.startswith("@"):
output_stream.write(line)
line = input_stream.readline()
n_1_read=Sam_record(line)
line = input_stream.readline()
n_2_read=Sam_record(line)
#We need three read in a row to assign one to the others
for line in input_stream:
read=Sam_record(line)
if n_1_read and n_2_read and read.get_query_name() == n_1_read.get_query_name() and \
read.get_query_name() == n_2_read.get_query_name():
three_reads={}
three_reads[test_read_for_assignation(read)]=read
three_reads[test_read_for_assignation(n_1_read)]=n_1_read
three_reads[test_read_for_assignation(n_2_read)]=n_2_read
#All 3 have been found and assigned
if not three_reads['first_assigned'].is_unmapped():
three_reads['first_unassigned'].set_reference_name(three_reads['first_assigned'].get_reference_name())
three_reads['first_unassigned'].set_unmapped_flag(False)
three_reads['first_unassigned'].set_position(three_reads['first_assigned'].get_position())
three_reads['first_unassigned'].set_cigar_string("%sM"%len(three_reads['first_unassigned'].get_query_sequence()))
three_reads['second'].set_reference_name(three_reads['first_assigned'].get_reference_name())
three_reads['second'].set_unmapped_flag(False)
three_reads['second'].set_position(three_reads['first_assigned'].get_position())
three_reads['second'].set_cigar_string("%sM"%len(three_reads['second'].get_query_sequence()))
output_stream.write(str(three_reads['first_unassigned']))
output_stream.write(str(three_reads['second']))
n_1_read=None
n_2_read=None
elif n_1_read and n_2_read:
logging.warning('Missing pair for singleton %s: is this file sorted.'%(n_2_read.get_query_name()))
output_stream.write(str(n_2_read))
n_2_read=n_1_read
n_1_read=read
elif n_1_read:
n_2_read=n_1_read
n_1_read=read
else:
n_1_read=read
def load_from_sites_generator(stream):
all_unmatched_read1={}
all_unmatched_read2={}
count_line=0
for line in stream:
count_line+=1
if count_line%10000==0:
sys.stderr.write('%s %s %s\n'%(count_line, len(all_unmatched_read1), len(all_unmatched_read2)))
sam_record = Sam_record(line)
if sam_record.is_first_read():
sam_record_r1 = sam_record
sam_record_r2 = all_unmatched_read2.pop(sam_record.get_query_name(),None)
if not sam_record_r2:
all_unmatched_read1[sam_record.get_query_name()]=sam_record
else:
sam_record_r2 = sam_record
sam_record_r1 = all_unmatched_read1.pop(sam_record.get_query_name(),None)
if not sam_record_r1:
all_unmatched_read2[sam_record.get_query_name()]=sam_record
if sam_record_r1 and sam_record_r2:
yield ((sam_record_r1,sam_record_r2))
sam_record = Sam_record(line)
if sam_record.get_reference_name(
) != current_reference and not current_reference is None:
#process this consensus
if current_reference != '*':
nb_dups, nb_uniq = find_duplicates(first_reads, second_reads,
distance_threshold)
total_nb_uniqs += nb_uniq
total_nb_dups += nb_dups
nb_fragment += len(second_reads)
output_reads(output_stream, first_reads, second_reads)
first_reads = {}
second_reads = {}
if sam_record.is_second_read():
second_reads[sam_record.get_query_name()] = sam_record
else:
first_reads[sam_record.get_query_name()] = sam_record
nb_reference += 1
if nb_reference % 1000 == 0:
print "process %s consensus" % nb_reference
current_reference = sam_record.get_reference_name()
if sam_record.get_reference_name(
) != current_reference and not current_reference is None:
#process this consensus
if current_reference != '*':
nb_dups = find_duplicates(first_reads, second_reads,
distance_threshold)
total_nb_dups += nb_dups
nb_fragment += len(second_reads)
output_reads(output_stream, first_reads, second_reads)
input_stream,process_input = utils_commands.get_output_stream_from_command(command)
command ="%s view -bS - | %s sort - %s"%(samtools_bin, samtools_bin, output_bam_file)
logging.info(command)
output_stream,process_output= utils_commands.get_input_stream_from_command(command)
#get the header
line = input_stream.readline()
while line.startswith("@"):
output_stream.write(line)
line = input_stream.readline()
while line:
read1=Sam_record(line)
line = input_stream.readline()
read2=Sam_record(line)
if read1.get_query_name() == read2.get_query_name():
if read1.is_second_read() and read2.is_first_read():
tmp = read1
read1=read2
read2=tmp
read2.set_reference_name(read1.get_reference_name())
output_stream.write(str(read1))
output_stream.write(str(read2))
else:
logging.critical("bam file is not sorted by read name")
input_stream.close()
output_stream.close()
#os.remove(output_bam_file+'.bam')
return
line = input_stream.readline()
output_stream.write(line)
continue
sam_record = Sam_record(line)
if sam_record.get_reference_name()!=current_reference and not current_reference is None:
#process this consensus
if current_reference!='*':
nb_dups, nb_uniq = find_duplicates(first_reads,second_reads, distance_threshold)
total_nb_uniqs+=nb_uniq
total_nb_dups+=nb_dups
nb_fragment+=len(second_reads)
output_reads(output_stream, first_reads, second_reads)
first_reads={}
second_reads={}
if sam_record.is_second_read():
second_reads[sam_record.get_query_name()]=sam_record
else:
first_reads[sam_record.get_query_name()]=sam_record
nb_reference+=1
if nb_reference%1000==0:
print "process %s consensus"%nb_reference
current_reference = sam_record.get_reference_name()
if sam_record.get_reference_name()!=current_reference and not current_reference is None:
#process this consensus
if current_reference!='*':
nb_dups = find_duplicates(first_reads,second_reads, distance_threshold)
total_nb_dups+=nb_dups
nb_fragment+=len(second_reads)
output_reads(output_stream, first_reads, second_reads)
library_size = estimate_library_size(nb_fragment, total_nb_uniqs)
print "%s fragments"%(nb_fragment)
