def checking(args, params, filenames):
log.logger.debug('started.')
try:
read_num_limit = params.quick_check_read_num
no_hhv_threshold = 3 / 1000000
need_check_threshold = 20 / 1000000
dr_threshold = 150 / 1000000
n_false = 0
n_need_check = 0
n_dr = 0
n_full = 0
finalfile = open(filenames.final_result, 'w')
finalfile.write(
'#file\tnum_unmapped_read_analyzed\tnum_read_mapped_to_HHV6\tHHV6_exists?\n'
)
for f in filenames.fpaths:
if args.file_type == 'rb':
infile = pysam.AlignmentFile(f, 'rb', check_sq=False)
elif args.file_type == 'rc':
infile = pysam.AlignmentFile(f,
'rc',
reference_filename=args.fa)
n = 0
with open(filenames.unmapped, 'w') as outfile:
tmp = []
for read in infile.fetch('*', until_eof=True):
if read.is_unmapped:
if not 'TAACCC' in read.query_sequence and not 'GGGTTA' in read.query_sequence:
if read.is_read1 is True:
header = '@%s/1' % read.query_name
else:
header = '@%s/2' % read.query_name
tmp.append(
'%s\n%s\n+\n%s\n' %
(header, read.query_sequence, read.qual))
n += 1
if len(tmp) == 100_000:
outfile.write(''.join(tmp))
tmp = []
if n == read_num_limit:
break
if len(tmp) >= 1:
outfile.write(''.join(tmp))
outfile.flush()
os.fdatasync(outfile.fileno())
infile.close()
if n == 0:
log.logger.info(
'No unmapped reads found in %s. Will continue anyway.' %
(n, f))
finalfile.write('%s\t%d\tNA\tNA\n' % (f, n))
utils.gzip_or_del(args, params, filenames.unmapped)
continue
elif n < read_num_limit:
log.logger.warning(
'Only %d unmapped reads were found in %s. Will continue anyway.'
% (n, f))
# mapping
cmd = 'hisat2 --mp %s -t -x %s -p %d -U %s --no-spliced-alignment > %s' % (
params.hisat2_mismatch_penalties, args.vrefindex, args.p,
filenames.unmapped, filenames.mapped_sam)
out = subprocess.run(cmd, shell=True, stderr=subprocess.PIPE)
log.logger.debug(
'\n' + '\n'.join([l.decode()
for l in out.stderr.splitlines()]))
if not out.returncode == 0:
log.logger.error('Error occurred during mapping.')
exit(1)
utils.gzip_or_del(args, params, filenames.unmapped)
# count mapped
mapped_n = 0
with open(filenames.mapped_sam) as infile:
for line in infile:
if not line[0] == '@':
ls = line.split()
if not ls[5] == '*':
readlen = len(ls[9])
if ls[5] == '%dM' % readlen:
mapped_n += 1
mapped_ratio = mapped_n / n
if mapped_ratio < no_hhv_threshold:
judge = 'False'
n_false += 1
elif mapped_ratio < need_check_threshold:
judge = 'Need_further_check'
n_need_check += 1
elif mapped_ratio < dr_threshold:
judge = 'likely_solo-DR'
n_dr += 1
else:
judge = 'likely_Full-length'
n_full += 1
finalfile.write('%s\t%d\t%d\t%s\n' % (f, n, mapped_n, judge))
utils.gzip_or_del(args, params, filenames.mapped_sam)
finalfile.flush()
os.fdatasync(finalfile.fileno())
log.logger.info(
'\n\n\033[34mQuick check result:\n\n No HHV-6 = %d\n Need check = %d\n Likely solo-DR = %d\n Likely Full-length = %d\033[0m\n\n \033[31mCaveats: This result is estimation and only for a screening purpose. This is not a conclusive result.\033[0m\n'
% (n_false, n_need_check, n_dr, n_full))
except:
log.logger.error('\n' + traceback.format_exc())
exit(1)
elif args.fastqin is True:
log.logger.info(
'Unmapped read retrieval skipped. Read1=%s, read2=%s.' %
(args.fq1, args.fq2))
if args.single is False:
filenames.unmapped_merged_1 = args.fq1
filenames.unmapped_merged_2 = args.fq2
else:
filenames.unmapped_merged_1 = args.fq1
# 1. mapping
import mapping
log.logger.info('Mapping of unmapped reads started.')
mapping.map_to_viruses(args, params, filenames)
if args.alignmentin is True:
utils.gzip_or_del(args, params, filenames.unmapped_merged_1)
utils.gzip_or_del(args, params, filenames.unmapped_merged_2)
if (args.ONT_bamin is False
and mapping.read_mapped is True) or args.ONT_bamin is True:
if args.ONT_bamin is True:
import mapping
filenames.mapped_to_virus_bam = args.ONT_bam
if args.remove_chr_with_no_read is True:
log.logger.info('Removing chrs without reads.')
mapping.remove_chrs_no_read(args, params, filenames, hhv6a_refid,
hhv6b_refid)
log.logger.info('BAM to bedgraph conversion started.')
mapping.bam_to_bedgraph(args, params, filenames)
# 2. identify high coverage viruses
def retrieve_unmapped_reads(args, params, filenames):
log.logger.debug('started.')
try:
if args.p <= 2:
thread_n = args.p
elif args.p >= 3:
thread_n = args.p - 1
# retrieve discordant reads, default
if args.use_mate_mapped is False and args.all_discordant is False:
if not args.b is None:
pysam.view('-@',
'%d' % thread_n,
'-f',
'12',
'-F',
'3842',
'-b',
'-o',
filenames.discordant_bam,
args.b,
catch_stdout=False)
elif not args.c is None:
pysam.view('-@',
'%d' % thread_n,
'-f',
'12',
'-F',
'3842',
'-b',
'-o',
filenames.discordant_bam,
'--reference',
args.fa,
args.c,
catch_stdout=False)
pysam.fastq('-@', '%d' % thread_n, '-N', '-0', '/dev/null', '-1',
filenames.unmapped_merged_pre1, '-2',
filenames.unmapped_merged_pre2, '-s', '/dev/null',
filenames.discordant_bam)
if args.keep is False:
os.remove(filenames.discordant_bam)
# retrieve discordant reads, non-default
else:
if not args.b is None:
pysam.view('-@',
'%d' % thread_n,
'-f',
'1',
'-F',
'3842',
'-b',
'-o',
filenames.discordant_bam,
args.b,
catch_stdout=False)
elif not args.c is None:
pysam.view('-@',
'%d' % thread_n,
'-f',
'1',
'-F',
'3842',
'-b',
'-o',
filenames.discordant_bam,
'--reference',
args.fa,
args.c,
catch_stdout=False)
pysam.sort('-@', '%d' % thread_n, '-n', '-O', 'BAM', '-o',
filenames.discordant_sort_bam, filenames.discordant_bam)
if args.keep is False:
os.remove(filenames.discordant_bam)
if args.all_discordant is True:
pysam.fastq('-@', '%d' % thread_n, '-N', '-0', '/dev/null',
'-1', filenames.unmapped_merged_pre1, '-2',
filenames.unmapped_merged_pre2, '-s', '/dev/null',
filenames.discordant_sort_bam)
else:
pysam.fastq('-@', '%d' % thread_n, '-f', '12', '-F', '3328',
'-N', '-0', '/dev/null', '-1',
filenames.unmapped_1, '-2', filenames.unmapped_2,
'-s', '/dev/null', filenames.discordant_sort_bam)
if args.use_mate_mapped is True:
pysam.view('-@',
'%d' % thread_n,
'-f',
'8',
'-F',
'3332',
'-b',
'-o',
filenames.unmapped_bam_3,
filenames.discordant_sort_bam,
catch_stdout=False)
pysam.view('-@',
'%d' % thread_n,
'-f',
'4',
'-F',
'3336',
'-b',
'-o',
filenames.unmapped_bam_4,
filenames.discordant_sort_bam,
catch_stdout=False)
pysam.merge('-@', '%d' % thread_n, '-f',
filenames.unmapped_bam_34,
filenames.unmapped_bam_3,
filenames.unmapped_bam_4)
pysam.sort('-@', '%d' % thread_n, '-n', '-O', 'BAM', '-o',
filenames.unmapped_sorted_34,
filenames.unmapped_bam_34)
pysam.fastq('-@', '%d' % thread_n, '-N', '-0', '/dev/null',
'-1', filenames.unmapped_3, '-2',
filenames.unmapped_4, '-s', '/dev/null',
filenames.unmapped_sorted_34)
# concatenate fastq
with open(filenames.unmapped_merged_pre1, 'w') as outfile:
for f in [filenames.unmapped_1, filenames.unmapped_3]:
if os.path.exists(f) is True:
with open(f) as infile:
for line in infile:
outfile.write(line)
utils.gzip_or_del(args, params, f)
with open(filenames.unmapped_merged_pre2, 'w') as outfile:
for f in [filenames.unmapped_2, filenames.unmapped_4]:
if os.path.exists(f) is True:
with open(f) as infile:
for line in infile:
outfile.write(line)
utils.gzip_or_del(args, params, f)
# remove short reads
infile1 = open(filenames.unmapped_merged_pre1)
infile2 = open(filenames.unmapped_merged_pre2)
outfile1 = open(filenames.unmapped_merged_1, 'w')
outfile2 = open(filenames.unmapped_merged_2, 'w')
min_seq_len = params.min_seq_len
tmp1, tmp2 = [], []
for line1, line2 in zip(infile1, infile2):
tmp1.append(line1)
tmp2.append(line2)
if len(tmp1) == 4:
seqlen1 = len(tmp1[1].strip())
seqlen2 = len(tmp2[1].strip())
if seqlen1 >= min_seq_len and seqlen2 >= min_seq_len:
outfile1.write(''.join(tmp1))
outfile2.write(''.join(tmp2))
tmp1, tmp2 = [], []
infile1.close()
infile2.close()
outfile1.close()
outfile2.close()
utils.gzip_or_del(args, params, filenames.unmapped_merged_pre1)
utils.gzip_or_del(args, params, filenames.unmapped_merged_pre2)
if args.keep is False:
if os.path.exists(filenames.discordant_sort_bam) is True:
os.remove(filenames.discordant_sort_bam)
if args.use_mate_mapped is True:
os.remove(filenames.unmapped_bam_3)
os.remove(filenames.unmapped_bam_4)
os.remove(filenames.unmapped_bam_34)
os.remove(filenames.unmapped_sorted_34)
except:
log.logger.error('\n' + traceback.format_exc())
exit(1)