def trim_reads(data_folder,
adaID,
VERBOSE=0,
summary=True,
quality=25,
blocksize=10,
minlen_read1=100,
minlen_read2=50):
'''Trim low quality at the end of reads'''
fn_in = get_read_filenames(data_folder, adaID, gzip=True)
fn_out = get_read_filenames(data_folder, adaID, gzip=True, trimmed=True)
n_good = 0
n_discarded = 0
with gzip.open(fn_in[0], 'rb') as fin1, \
gzip.open(fn_in[1], 'rb') as fin2, \
gzip.open(fn_out[0], 'rb') as fout1, \
gzip.open(fn_out[1], 'wb') as fout2:
it1 = SeqIO.read(fin1, 'fastq')
it2 = SeqIO.read(fin2, 'fastq')
for irp, reads in enumerate(izip(it1, it2)):
if VERBOSE >= 2:
if not ((irp + 1) % 10000):
print irp + 1
# Trim both reads
trims = [
trim_read(read, quality=quality, blocksize=blocksize)
for read in reads
]
lrs = map(len, trims)
if (lrs[0] > minlen_read1) and (lrs[1] > minlen_read2):
SeqIO.write(trims[0], fout1, 'fastq')
SeqIO.write(trims[1], fout2, 'fastq')
n_good += 1
else:
n_discarded += 1
if VERBOSE:
print 'Trim lowq ends of reads:'
print 'Good:', n_good
print 'Discarded:', n_discarded
# Write summary to file
if summary:
with open(get_trim_summary_filename(data_folder, adaID), 'a') as f:
f.write('\n')
f.write('Trim low quality ends results: adaID ' + adaID + '\n')
f.write('Total:\t\t' + str(irp) + '\n')
f.write('Good:\t\t' + str(n_good) + '\n')
f.write('Discarded:\t' + str(n_discarded) + '\n')
def trim_reads(data_folder, adaID, VERBOSE=0, summary=True, quality=25, blocksize=10,
minlen_read1=100, minlen_read2=50):
'''Trim low quality at the end of reads'''
fn_in = get_read_filenames(data_folder, adaID, gzip=True)
fn_out = get_read_filenames(data_folder, adaID, gzip=True, trimmed=True)
n_good = 0
n_discarded = 0
with gzip.open(fn_in[0], 'rb') as fin1, \
gzip.open(fn_in[1], 'rb') as fin2, \
gzip.open(fn_out[0], 'rb') as fout1, \
gzip.open(fn_out[1], 'wb') as fout2:
it1 = SeqIO.read(fin1, 'fastq')
it2 = SeqIO.read(fin2, 'fastq')
for irp, reads in enumerate(izip(it1, it2)):
if VERBOSE >= 2:
if not ((irp + 1) % 10000):
print irp + 1
# Trim both reads
trims = [trim_read(read, quality=quality, blocksize=blocksize)
for read in reads]
lrs = map(len, trims)
if (lrs[0] > minlen_read1) and (lrs[1] > minlen_read2):
SeqIO.write(trims[0], fout1, 'fastq')
SeqIO.write(trims[1], fout2, 'fastq')
n_good += 1
else:
n_discarded += 1
if VERBOSE:
print 'Trim lowq ends of reads:'
print 'Good:', n_good
print 'Discarded:', n_discarded
# Write summary to file
if summary:
with open(get_trim_summary_filename(data_folder, adaID), 'a') as f:
f.write('\n')
f.write('Trim low quality ends results: adaID '+adaID+'\n')
f.write('Total:\t\t'+str(irp)+'\n')
f.write('Good:\t\t'+str(n_good)+'\n')
f.write('Discarded:\t'+str(n_discarded)+'\n')
def gunzip_demultiplexed_reads(data_folder, adaID, VERBOSE=0):
'''Gunzip FastQ.gz demultiplexed files'''
from hivwholeseq.sequencing.filenames import get_read_filenames
fns = get_read_filenames(data_folder, adaID, gzip=True)
for fn in fns:
if not os.path.isfile(fn):
continue
sp.call(['gunzip', fn])
if VERBOSE >= 2:
print 'Gunzipped:', fn
def premap_stampy(data_folder,
adaID,
VERBOSE=0,
threads=1,
summary=True,
maxreads=-1,
subsrate=0.05,
gapopen=40,
gapextend=3):
'''Call stampy for actual mapping'''
if VERBOSE:
print 'Premapping: adaID ', adaID
if summary:
summary_filename = get_premap_summary_filename(data_folder, adaID)
# Stampy can handle both gzipped and uncompressed fastq inputs
input_filenames = get_read_filenames(data_folder, adaID, gzip=True)
if not os.path.isfile(input_filenames[0]):
input_filenames = get_read_filenames(data_folder, adaID, gzip=False)
if not all(map(os.path.isfile, input_filenames)):
raise OSError('Input files for mapping not found: ' +
input_filenames[0])
# parallelize if requested
if threads == 1:
call_list = [
stampy_bin,
'--overwrite',
'-g',
get_reference_premap_index_filename(data_folder, adaID, ext=False),
'-h',
get_reference_premap_hash_filename(data_folder, adaID, ext=False),
'-o',
get_premapped_filename(data_folder, adaID, type='sam'),
'--insertsize=450',
'--insertsd=100',
'--substitutionrate=' + str(subsrate),
'--gapopen=' + str(gapopen),
'--gapextend=' + str(gapextend),
]
if maxreads > 0:
call_list.append('--numrecords=' + str(maxreads))
call_list.extend(['-M'] + input_filenames)
call_list = map(str, call_list)
if VERBOSE >= 2:
print ' '.join(call_list)
sp.call(call_list)
if summary:
with open(get_premap_summary_filename(data_folder, adaID),
'a') as f:
f.write('\nStampy premapped (single thread).\n')
# Convert to compressed BAM
convert_sam_to_bam(
get_premapped_filename(data_folder, adaID, type='bam'))
if summary:
with open(summary_filename, 'a') as f:
f.write('\nSAM file converted to compressed BAM: '+\
get_premapped_filename(data_folder, adaID, type='bam')+'\n')
else:
# Multithreading works as follows: call qsub + stampy, monitor the process
# IDs with qstat at regular intervals, and finally merge results with pysam
output_file_parts = [
get_premapped_filename(data_folder,
adaID,
type='bam',
part=(j + 1)) for j in xrange(threads)
]
# Submit map script
jobs_done = np.zeros(threads, bool)
job_IDs = np.zeros(threads, 'S30')
# Submit map call
import hivwholeseq
JOBDIR = hivwholeseq.__path__[0].rstrip('/') + '/'
JOBLOGOUT = JOBDIR + 'logout'
JOBLOGERR = JOBDIR + 'logerr'
cluster_time = ['23:59:59', '1:59:59']
vmem = '8G'
for j in xrange(threads):
call_list = [
'qsub', '-cwd', '-b', 'y', '-S', '/bin/bash', '-o', JOBLOGOUT,
'-e', JOBLOGERR, '-N', adaID + ' p' + str(j + 1), '-l',
'h_rt=' + cluster_time[threads >= 30], '-l', 'h_vmem=' + vmem,
stampy_bin, '--overwrite', '-g',
get_reference_premap_index_filename(
data_folder, adaID, ext=False), '-h',
get_reference_premap_hash_filename(
data_folder, adaID, ext=False), '-o',
get_premapped_filename(
data_folder, adaID, type='sam', part=(j + 1)),
'--processpart=' + str(j + 1) + '/' + str(threads),
'--insertsize=450', '--insertsd=100', '--substitutionrate=' +
str(subsrate), '--gapopen=' + str(gapopen),
'--gapextend=' + str(gapextend), '-M'
] + input_filenames
call_list = map(str, call_list)
if VERBOSE >= 2:
print ' '.join(call_list)
job_ID = sp.check_output(call_list)
job_ID = job_ID.split()[2]
job_IDs[j] = job_ID
# Monitor output
time_wait = 10 # secs
while not jobs_done.all():
# Sleep some time
time.sleep(time_wait)
# Get the output of qstat to check the status of jobs
qstat_output = sp.check_output(['qstat'])
qstat_output = qstat_output.split(
'\n')[:-1] # The last is an empty line
if VERBOSE >= 3:
print qstat_output
if len(qstat_output) < 3:
jobs_done[:] = True
break
else:
qstat_output = [line.split()[0] for line in qstat_output[2:]]
time_wait = 10 # secs
for j in xrange(threads):
if jobs_done[j]:
continue
if job_IDs[j] not in qstat_output:
# Convert to BAM for merging
if VERBOSE >= 1:
print 'Convert premapped reads to BAM for merging: adaID '+\
adaID+', part '+str(j+1)+ ' of '+ \
str(threads)
convert_sam_to_bam(output_file_parts[j])
# We do not need to wait if we did the conversion (it takes
# longer than some secs)
time_wait = 0
jobs_done[j] = True
if summary:
with open(summary_filename, 'a') as f:
f.write('Stampy premapped (' + str(threads) + ' threads).\n')
# Concatenate output files
if VERBOSE >= 1:
print 'Concatenate premapped reads: adaID ' + adaID + '...',
output_filename = get_premapped_filename(data_folder,
adaID,
type='bam',
unsorted=True)
pysam.cat('-o', output_filename, *output_file_parts)
if VERBOSE >= 1:
print 'done.'
if summary:
with open(summary_filename, 'a') as f:
f.write('BAM files concatenated (unsorted).\n')
# Sort the file by read names (to ensure the pair_generator)
# NOTE: we exclude the extension and the option -f because of a bug in samtools
if VERBOSE >= 1:
print 'Sort premapped reads: adaID ' + adaID
output_filename_sorted = get_premapped_filename(data_folder,
adaID,
type='bam',
unsorted=False)
pysam.sort('-n', output_filename, output_filename_sorted[:-4])
if summary:
with open(summary_filename, 'a') as f:
f.write('Joint BAM file sorted.\n')
# Reheader the file without BAM -> SAM -> BAM
if VERBOSE >= 1:
print 'Reheader premapped reads: adaID ' + adaID
header_filename = get_premapped_filename(data_folder,
adaID,
type='sam',
part=1)
pysam.reheader(header_filename, output_filename_sorted)
if summary:
with open(summary_filename, 'a') as f:
f.write('Joint BAM file reheaded.\n')
if VERBOSE >= 1:
print 'Remove temporary files: adaID ' + adaID
remove_premapped_tempfiles(data_folder, adaID, VERBOSE=VERBOSE)
if summary:
with open(summary_filename, 'a') as f:
f.write('Temp premapping files removed.\n')
f.write('\n')
parser.add_argument('--adaID',
required=True,
help='Adapter ID to analyze (e.g. TS2)')
args = parser.parse_args()
seq_run = args.run
VERBOSE = args.verbose
maxreads = args.maxreads
adaID = args.adaID
# Specify the dataset
dataset = MiSeq_runs[seq_run]
data_folder = dataset['folder']
# Get some reads
fns = get_read_filenames(data_folder, adaID, gzip=True)
with gzip.open(fns[0], 'rb') as fh1, gzip.open(fns[1], 'rb') as fh2:
reads_iter1 = SeqIO.parse(fh1, 'fastq')
reads_iter2 = SeqIO.parse(fh2, 'fastq')
read_pairs = []
inds = 20000 + np.arange(100000)
np.random.shuffle(inds)
inds = np.sort(inds[:10])
ii = 0
for irp, reads in enumerate(izip(reads_iter1, reads_iter2)):
if irp == inds[ii]:
read_pairs.append(reads)
ii += 1
if ii == len(inds):
def demultiplex_reads_single_index(data_folder,
data_filenames,
adapters_designed,
maxreads=-1,
VERBOSE=0,
summary=True):
'''Demultiplex reads with single index adapters'''
# Get the read filenames
datafile_read1 = data_filenames['read1']
datafile_read2 = data_filenames['read2']
datafile_adapter = data_filenames['adapter']
# Open output files (compressed)
fouts = {
adaID: [
gzip.open(fn, 'wb', compresslevel=9)
for fn in get_read_filenames(data_folder, adaID, gzip=True)
]
for adaID, _ in adapters_designed
}
fouts['unclassified'] = [
gzip.open(fn, 'wb', compresslevel=9)
for fn in get_unclassified_reads_filenames(data_folder, gzip=True)
]
adapters_designed_inv = dict(map(reversed, adapters_designed))
adapters_strings = map(itemgetter(1), adapters_designed)
# Make sure you close the files
try:
# Iterate over all reads (using fast iterators)
with gzip.open(datafile_read1, 'rb') as fh1,\
gzip.open(datafile_read2, 'rb') as fh2,\
gzip.open(datafile_adapter, 'rb') as fha:
if VERBOSE >= 3:
print 'adaID'
print '--------------------'
adapters_found = Counter()
for i, (read1, read2, adapter) in enumerate(
izip(FGI(fh1), FGI(fh2), SeqIO.parse(fha, 'fastq'))):
if i == maxreads:
if VERBOSE:
print 'Maxreads reached.'
break
# Print some output
if VERBOSE and (not ((i + 1) % 10000)):
print i + 1
# If the adapter is not known, add it to the list
adapter_string = str(adapter.seq)
adapters_found[adapter_string] += 1
# If the adapter does not match any know one,
# throw into wastebin folder
if adapter_string not in adapters_strings:
adaID = 'unclassified'
else:
adaID = adapters_designed_inv[adapter_string]
if VERBOSE >= 3:
print adaID
# Write sequences (append to file, manual but fast)
fouts[adaID][0].write("@%s\n%s\n+\n%s\n" % read1)
fouts[adaID][1].write("@%s\n%s\n+\n%s\n" % read2)
if adapter_string not in adapters_strings:
SeqIO.write(adapter, fouts['unclassified'][2], 'fastq')
finally:
# Close all adaIDs
for fout in fouts.itervalues():
# Close both read 1 and read 2 (and barcode for unclassified)
for fou in fout:
fou.close()
if summary:
with open(get_demultiplex_summary_filename(data_folder), 'a') as f:
f.write('\n')
f.write('Total number of reads demultiplexed: ' + str(i + 1) +
'\n')
f.write('Adapters found across all reads:\n')
for e in adapters_found.most_common():
f.write('\t'.join(map(str, e)) + '\n')
seq_run = args.run
VERBOSE = args.verbose
submit = args.submit
maxreads = args.maxreads
adaID = args.adaID
savefig = args.savefig
if submit:
fork_self(seq_run, VERBOSE=VERBOSE, maxreads=maxreads, savefig=savefig)
sys.exit()
dataset = load_sequencing_run(seq_run)
data_folder = dataset.folder
read_len = dataset.cycles // 2
reads_filenames = get_read_filenames(data_folder, adaID, gzip=True)
if not os.path.isfile(reads_filenames[0]):
reads_filenames = get_read_filenames(data_folder, adaID, gzip=False)
title = seq_run + ", " + adaID
quality = quality_score_along_reads(
read_len, reads_filenames, randomreads=(maxreads >= 1), maxreads=maxreads, VERBOSE=VERBOSE
)
plot_cuts_quality_along_reads(data_folder, adaID, quality, title=title, VERBOSE=VERBOSE, savefig=savefig)
# if plotfull:
# plot_quality_along_reads(data_folder, adaID, title,
# quality, VERBOSE=VERBOSE,
# savefig=savefig)
help='Maximal number of reads to analyze')
parser.add_argument('--adaID', required=True,
help='Adapter ID to analyze (e.g. TS2)')
args = parser.parse_args()
seq_run = args.run
VERBOSE = args.verbose
maxreads = args.maxreads
adaID = args.adaID
# Specify the dataset
dataset = MiSeq_runs[seq_run]
data_folder = dataset['folder']
# Get some reads
fns = get_read_filenames(data_folder, adaID, gzip=True)
with gzip.open(fns[0], 'rb') as fh1, gzip.open(fns[1], 'rb') as fh2:
reads_iter1 = SeqIO.parse(fh1, 'fastq')
reads_iter2 = SeqIO.parse(fh2, 'fastq')
read_pairs = []
inds = 20000 + np.arange(100000); np.random.shuffle(inds); inds = np.sort(inds[:10])
ii = 0
for irp, reads in enumerate(izip(reads_iter1, reads_iter2)):
if irp == inds[ii]:
read_pairs.append(reads)
ii += 1
if ii == len(inds):
break
def premap_stampy(data_folder,
adaID,
VERBOSE=0,
threads=1,
summary=True,
maxreads=-1,
subsrate=0.05,
gapopen=40,
gapextend=3):
'''Call stampy for actual mapping'''
if VERBOSE:
print 'Premapping: adaID ', adaID
if summary:
summary_filename = get_premap_summary_filename(data_folder, adaID)
# Stampy can handle both gzipped and uncompressed fastq inputs
input_filenames = get_read_filenames(data_folder, adaID, gzip=True)
if not os.path.isfile(input_filenames[0]):
input_filenames = get_read_filenames(data_folder, adaID, gzip=False)
if not all(map(os.path.isfile, input_filenames)):
raise OSError('Input files for mapping not found: ' +
input_filenames[0])
# parallelize if requested
if threads == 1:
call_list = [
stampy_bin,
'--overwrite',
'-g',
get_reference_premap_index_filename(data_folder, adaID, ext=False),
'-h',
get_reference_premap_hash_filename(data_folder, adaID, ext=False),
'-o',
get_premapped_filename(data_folder, adaID, type='sam'),
'--insertsize=450',
'--insertsd=100',
'--substitutionrate=' + str(subsrate),
'--gapopen=' + str(gapopen),
'--gapextend=' + str(gapextend),
]
if maxreads > 0:
call_list.append('--numrecords=' + str(maxreads))
call_list.extend(['-M'] + input_filenames)
call_list = map(str, call_list)
if VERBOSE >= 2:
print ' '.join(call_list)
sp.call(call_list)
if summary:
with open(get_premap_summary_filename(data_folder, adaID),
'a') as f:
f.write('\nStampy premapped (single thread).\n')
# Convert to compressed BAM
convert_sam_to_bam(
get_premapped_filename(data_folder, adaID, type='bam'))
if summary:
with open(summary_filename, 'a') as f:
f.write('\nSAM file converted to compressed BAM: '+\
get_premapped_filename(data_folder, adaID, type='bam')+'\n')
else:
# Multithreading works as follows: call qsub + stampy, monitor the process
# IDs with qstat at regular intervals, and finally merge results with pysam
output_file_parts = [
get_premapped_filename(
data_folder, adaID, type='bam', part=(j + 1))
for j in xrange(threads)
]
# Submit map script
jobs_done = np.zeros(threads, bool)
job_IDs = np.zeros(threads, 'S30')
# Submit map call
import hivwholeseq
JOBDIR = hivwholeseq.__path__[0].rstrip('/') + '/'
JOBLOGOUT = JOBDIR + 'logout'
JOBLOGERR = JOBDIR + 'logerr'
cluster_time = ['23:59:59', '1:59:59']
vmem = '8G'
for j in xrange(threads):
call_list = [
'qsub', '-cwd', '-b', 'y', '-S', '/bin/bash', '-o', JOBLOGOUT,
'-e', JOBLOGERR, '-N', adaID + ' p' + str(j + 1), '-l',
'h_rt=' + cluster_time[threads >= 30], '-l', 'h_vmem=' + vmem,
stampy_bin, '--overwrite', '-g',
get_reference_premap_index_filename(
data_folder, adaID, ext=False), '-h',
get_reference_premap_hash_filename(
data_folder, adaID, ext=False), '-o',
get_premapped_filename(
data_folder, adaID, type='sam', part=(j + 1)),
'--processpart=' + str(j + 1) + '/' + str(threads),
'--insertsize=450', '--insertsd=100', '--substitutionrate=' +
str(subsrate), '--gapopen=' + str(gapopen),
'--gapextend=' + str(gapextend), '-M'
] + input_filenames
call_list = map(str, call_list)
if VERBOSE >= 2:
print ' '.join(call_list)
job_ID = sp.check_output(call_list)
job_ID = job_ID.split()[2]
job_IDs[j] = job_ID
# Monitor output
time_wait = 10 # secs
while not jobs_done.all():
# Sleep some time
time.sleep(time_wait)
# Get the output of qstat to check the status of jobs
qstat_output = sp.check_output(['qstat'])
qstat_output = qstat_output.split(
'\n')[:-1] # The last is an empty line
if VERBOSE >= 3:
print qstat_output
if len(qstat_output) < 3:
jobs_done[:] = True
break
else:
qstat_output = [line.split()[0] for line in qstat_output[2:]]
time_wait = 10 # secs
for j in xrange(threads):
if jobs_done[j]:
continue
if job_IDs[j] not in qstat_output:
# Convert to BAM for merging
if VERBOSE >= 1:
print 'Convert premapped reads to BAM for merging: adaID '+\
adaID+', part '+str(j+1)+ ' of '+ \
str(threads)
convert_sam_to_bam(output_file_parts[j])
# We do not need to wait if we did the conversion (it takes
# longer than some secs)
time_wait = 0
jobs_done[j] = True
if summary:
with open(summary_filename, 'a') as f:
f.write('Stampy premapped (' + str(threads) + ' threads).\n')
# Concatenate output files
if VERBOSE >= 1:
print 'Concatenate premapped reads: adaID ' + adaID + '...',
output_filename = get_premapped_filename(
data_folder, adaID, type='bam', unsorted=True)
pysam.cat('-o', output_filename, *output_file_parts)
if VERBOSE >= 1:
print 'done.'
if summary:
with open(summary_filename, 'a') as f:
f.write('BAM files concatenated (unsorted).\n')
# Sort the file by read names (to ensure the pair_generator)
# NOTE: we exclude the extension and the option -f because of a bug in samtools
if VERBOSE >= 1:
print 'Sort premapped reads: adaID ' + adaID
output_filename_sorted = get_premapped_filename(
data_folder, adaID, type='bam', unsorted=False)
pysam.sort('-n', output_filename, output_filename_sorted[:-4])
if summary:
with open(summary_filename, 'a') as f:
f.write('Joint BAM file sorted.\n')
# Reheader the file without BAM -> SAM -> BAM
if VERBOSE >= 1:
print 'Reheader premapped reads: adaID ' + adaID
header_filename = get_premapped_filename(
data_folder, adaID, type='sam', part=1)
pysam.reheader(header_filename, output_filename_sorted)
if summary:
with open(summary_filename, 'a') as f:
f.write('Joint BAM file reheaded.\n')
if VERBOSE >= 1:
print 'Remove temporary files: adaID ' + adaID
remove_premapped_tempfiles(data_folder, adaID, VERBOSE=VERBOSE)
if summary:
with open(summary_filename, 'a') as f:
f.write('Temp premapping files removed.\n')
f.write('\n')
if VERBOSE >= 3:
print 'adaIDs', adaIDs
# Iterate over adaIDs
for adaID in adaIDs:
outdir = os.getcwd()+'/q_control_'+adaID
if not os.path.exists(outdir):
try:
os.mkdir(outdir)
except:
print "cannot make directory:",outdir
sys.exit(1)
# Repeat both for read 1 and read 2
datafiles = get_read_filenames(data_folder, adaID, filtered=False)
datafiles = {'read 1': datafiles[0],
'read 2': datafiles[1]}
histograms = {}
for readname, datafile in datafiles.iteritems():
# Result data structures
longest_good_block = np.zeros(L+1, int)
first_bad_nucleotide = np.zeros(L+1, int)
phred_score_dis = np.zeros((50,L))
# Read data
with open(datafile, 'r') as f:
seq_iter = SeqIO.parse(f, 'fastq')
def demultiplex_reads_single_index(data_folder, data_filenames, adapters_designed,
maxreads=-1, VERBOSE=0, summary=True):
'''Demultiplex reads with single index adapters'''
# Get the read filenames
datafile_read1 = data_filenames['read1']
datafile_read2 = data_filenames['read2']
datafile_adapter = data_filenames['adapter']
# Open output files (compressed)
fouts = {adaID: [gzip.open(fn, 'wb', compresslevel=9)
for fn in get_read_filenames(data_folder, adaID, gzip=True)]
for adaID, _ in adapters_designed}
fouts['unclassified'] = [gzip.open(fn, 'wb', compresslevel=9)
for fn in get_unclassified_reads_filenames(data_folder, gzip=True)]
adapters_designed_inv = dict(map(reversed, adapters_designed))
adapters_strings = map(itemgetter(1), adapters_designed)
# Make sure you close the files
try:
# Iterate over all reads (using fast iterators)
with gzip.open(datafile_read1, 'rb') as fh1,\
gzip.open(datafile_read2, 'rb') as fh2,\
gzip.open(datafile_adapter, 'rb') as fha:
if VERBOSE >= 3:
print 'adaID'
print '--------------------'
adapters_found = Counter()
for i, (read1, read2, adapter) in enumerate(izip(FGI(fh1), FGI(fh2),
SeqIO.parse(fha, 'fastq'))):
if i == maxreads:
if VERBOSE:
print 'Maxreads reached.'
break
# Print some output
if VERBOSE and (not ((i + 1) % 10000)):
print i + 1
# If the adapter is not known, add it to the list
adapter_string = str(adapter.seq)
adapters_found[adapter_string] += 1
# If the adapter does not match any know one,
# throw into wastebin folder
if adapter_string not in adapters_strings:
adaID = 'unclassified'
else:
adaID = adapters_designed_inv[adapter_string]
if VERBOSE >= 3:
print adaID
# Write sequences (append to file, manual but fast)
fouts[adaID][0].write("@%s\n%s\n+\n%s\n" % read1)
fouts[adaID][1].write("@%s\n%s\n+\n%s\n" % read2)
if adapter_string not in adapters_strings:
SeqIO.write(adapter, fouts['unclassified'][2], 'fastq')
finally:
# Close all adaIDs
for fout in fouts.itervalues():
# Close both read 1 and read 2 (and barcode for unclassified)
for fou in fout:
fou.close()
if summary:
with open(get_demultiplex_summary_filename(data_folder), 'a') as f:
f.write('\n')
f.write('Total number of reads demultiplexed: '+str(i+1)+'\n')
f.write('Adapters found across all reads:\n')
for e in adapters_found.most_common():
f.write('\t'.join(map(str, e))+'\n')
seq_run = args.run
VERBOSE = args.verbose
submit = args.submit
maxreads = args.maxreads
adaID = args.adaID
savefig = args.savefig
if submit:
fork_self(seq_run, VERBOSE=VERBOSE, maxreads=maxreads, savefig=savefig)
sys.exit()
dataset = load_sequencing_run(seq_run)
data_folder = dataset.folder
read_len = dataset.cycles // 2
reads_filenames = get_read_filenames(data_folder, adaID, gzip=True)
if not os.path.isfile(reads_filenames[0]):
reads_filenames = get_read_filenames(data_folder, adaID, gzip=False)
title = seq_run + ', ' + adaID
quality = quality_score_along_reads(read_len,
reads_filenames,
randomreads=(maxreads >= 1),
maxreads=maxreads,
VERBOSE=VERBOSE)
plot_cuts_quality_along_reads(data_folder,
adaID,
quality,
title=title,
VERBOSE=VERBOSE,
