def Loc_paired(read, loc_dict):
'''
Record the location information for paired sequencing
Return:
loc_s = { chrid: {loc1: coverage; ...}; chrid:..}
loc1: plus strand (++ or --) 5' pos + minus strand (+- or -+) 5' pos
'''
# Using the read class In: paired unique mapping read Out: [flag,strand,chrom,pos1,CIGAR,pos2,insert,seq,score]
read_s = RI(read)
read_info = read_s.extract_information()
if len(read_info) == 0:
return loc_dict
strand, chrom, pos, pos2, insert = read_info[1], read_info[2], read_info[
3], read_info[5], read_info[6]
# only use one mate with plus insert to record
if strand == '+-' or strand == '-+':
return loc_dict
site = pos + '_' + str(int(pos) + int(insert) - 1)
if loc_dict.has_key(chrom):
if loc_dict[chrom].has_key(site):
loc_dict[chrom][site] += 1
else:
loc_dict[chrom][site] = 1
else:
loc_dict[chrom] = {}
loc_dict[chrom][site] = 1
return loc_dict
def Loc_paired(read,loc_dict):
'''
Record the location information for paired sequencing
Return:
loc_s = { chrid: {loc1: coverage; ...}; chrid:..}
loc1: plus strand (++ or --) 5' pos + minus strand (+- or -+) 5' pos
'''
# Using the read class In: paired unique mapping read Out: [flag,strand,chrom,pos1,CIGAR,pos2,insert,seq,score]
read_s = RI(read)
read_info = read_s.extract_information()
if len(read_info) == 0:
return loc_dict
strand, chrom, pos, pos2, insert = read_info[1], read_info[2], read_info[3], read_info[5], read_info[6]
# only use one mate with plus insert to record
if strand == '+-' or strand == '-+':
return loc_dict
site = pos + '_' + str(int(pos) + int(insert) - 1)
if loc_dict.has_key(chrom):
if loc_dict[chrom].has_key(site):
loc_dict[chrom][site] +=1
else:
loc_dict[chrom][site] = 1
else:
loc_dict[chrom] = {}
loc_dict[chrom][site] = 1
return loc_dict
def Loc_single(read, loc_dict):
'''
Record the location information for single-end sequencing
Return:
loc_s = { chrid: {loc1: coverage; ...}; chrid:..}
loc1: 5' pos + strand
'''
# Using the read class In: single unique mapping read Out: [flag,strand,chrom,pos,CIGAR,seq,score]
read_s = RI(read)
read_info = read_s.extract_information()
if len(read_info) == 0:
return loc_dict
strand, chrom, pos, seq = read_info[1], read_info[2], read_info[
3], read_info[5]
# The minus strand should be shifted len(seq) bp
if strand == '-+':
pos = int(pos) + len(seq) - 1
site = strand + '_' + str(pos)
if loc_dict.has_key(chrom):
if loc_dict[chrom].has_key(site):
loc_dict[chrom][site] += 1
else:
loc_dict[chrom][site] = 1
else:
loc_dict[chrom] = {}
loc_dict[chrom][site] = 1
return loc_dict
def Loc_single(read,loc_dict):
'''
Record the location information for single-end sequencing
Return:
loc_s = { chrid: {loc1: coverage; ...}; chrid:..}
loc1: 5' pos + strand
'''
# Using the read class In: single unique mapping read Out: [flag,strand,chrom,pos,CIGAR,seq,score]
read_s = RI(read)
read_info = read_s.extract_information()
if len(read_info) == 0:
return loc_dict
strand, chrom, pos, seq = read_info[1], read_info[2], read_info[3], read_info[5]
# The minus strand should be shifted len(seq) bp
if strand == '-+':
pos = int(pos) + len(seq) -1
site = strand + '_' + str(pos)
if loc_dict.has_key(chrom):
if loc_dict[chrom].has_key(site):
loc_dict[chrom][site] += 1
else:
loc_dict[chrom][site] = 1
else:
loc_dict[chrom] = {}
loc_dict[chrom][site] = 1
return loc_dict
def get_read_info(self):
'''
Using the read class
If the read is unique mapping or unique and paired mapping, we will get a information list.
If not, we will get a empty list ([]).
In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
'''
read_s = RI(self.read)
read_info = read_s.extract_information()
return read_info
def get_read_info(self):
"""
Using the read class
If the read is unique mapping or unique and paired mapping, we will get a information list.
If not, we will get a empty list ([]).
In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
"""
read_s = RI(self.read, self.bsmb)
read_info = read_s.extract_information()
return read_info
def run(args):
"""
Alternative module: Use the strategy in Bis-SNP to trim 5' bisulfite conversion failures
"""
options = args.parse_args()
if len(options.sam_file) == 0:
error("Missing the SAM file, use -s or --sam option.")
else:
options.sam_file = options.sam_file.split(',')
for s in options.sam_file:
if not os.path.isfile(s):
error("Can't open the SAM file: " + s)
sys.exit(1)
if len(options.ref_file) == 0:
error(
"Missing the reference genome fasta file, use -r or --ref option.")
else:
if not os.path.isfile(options.ref_file):
error("Can't open the ref file: " + options.ref_file)
if len(options.samtools) != 0:
if options.samtools[-1] != '/':
options.samtools += '/'
if len(options.name) == 0:
error("Missing the output file name, use -n or --name options.")
sam_inf = options.sam_file
ref_file = options.ref_file
bsm = options.bsm
s_path = options.samtools
name = options.name
remove_overlap = options.remove_overlap
filter_dup = options.filter_dup
p_poisson = options.p_poisson
gsize = options.gsize
not_mapping = options.not_mapping
info("Get the all parameter!!")
#check the input mapping files
sam_format, read_inf = check.check_mapping_file_flag(sam_inf[0], s_path)
pre_flag = read_inf.readline().split('\t')[1]
if 'p' in pre_flag:
single_on = False
info("The input mapping files are paired-end sequencing!")
else:
single_on = True
info("The input mapping files are single-end sequencing!")
loc_dict = {}
if filter_dup:
## if filter_up is TRUE, the duplicate reads will be assessed and shown in Dup_dis.pdf
info("The filter_dup has been set True.")
info("Assess the duplicate reads...")
for sam in sam_inf:
#check the input mapping files
sam_format, read_inf = check.check_mapping_file(sam, s_path)
if single_on:
for read in read_inf:
loc_dict = LI.Loc_single(read, loc_dict, bsm)
else:
for read in read_inf:
loc_dict = LI.Loc_paired(read, loc_dict, bsm)
max_cov = DR.duplicate_report(loc_dict, gsize, p_poisson, name)
info('Get the duplicate reads distribution!')
#get reference information
ref = GR.get_ref(ref_file)
trim_position = []
filter_duplicate_reads = 0
filter_nonuniform_trim_bp = 0
filter_nonuniform_trim_bp_CG = 0
filter_remove_overlap_bp = 0
filter_not_mapping_reads = 0
all_reads = 0
not_mapping_reads = 0
all_mapping_bp = 0
##filter the 5' bisulfite failure
for sam in sam_inf:
out_sam = sam[:-4] + '_' + name + '_filter.sam'
out = open(out_sam, 'w')
#check the input mapping files
record_mate = {}
sam_format, read_inf = check.check_mapping_file_header(sam, s_path)
for read in read_inf:
#for sam header
if read.startswith('@'):
out.write(read)
continue
else:
all_reads += 1 ##record the read number (2013-06-20)
#Get the read information for trimming
#If the read isn't unique mapping, we will get a empty list ([]).
#In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
#In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
read_info = RI(read, bsm)
read_info = read_info.extract_information()
if len(read_info) == 0:
not_mapping_reads += 1
if not_mapping: #keep the not_unique mapping reads (or not paired mapping)
out.write(read)
else:
filter_not_mapping_reads += 1 ##record the not mapping read number (2013-06-20)
continue
if len(
loc_dict
) > 0: #the --filter_dup has been set True, have to remove duplicate reads
duplicate, loc_dict = DF(read_info, loc_dict, max_cov,
single_on)
else:
duplicate = False
if single_on:
all_mapping_bp += len(
read_info[5]
) ##record the mapping read basepair (2013-06-20)
else:
all_mapping_bp += len(
read_info[7]
) ##record the mapping read basepair (2013-06-20)
record_mate, trim_position, filter_nonuniform_trim_bp_CG, filter_duplicate_reads, filter_remove_overlap_bp = NF.nonuniform_filter(
read, out, read_info, ref, remove_overlap, duplicate,
single_on, record_mate, trim_position,
filter_nonuniform_trim_bp_CG, filter_duplicate_reads,
filter_remove_overlap_bp)
out.close()
del record_mate
NR.nonuniform_generator(trim_position, name)
for i in range(len(trim_position)):
filter_nonuniform_trim_bp += i * trim_position[i]
##produce the filter report
info('Produce the report file...')
report_out = open(name + "_BSeQC_nonuniform_filter_report.txt", 'w')
report_out.write('Total reads: %d\n' % all_reads)
if single_on:
report_out.write('Not unique mapping reads: %d(%.2f%s all reads)\n' %
(not_mapping_reads,
float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write(
'Unique mapping reads: %d(%.2f%s all reads)\n' %
((all_reads - not_mapping_reads),
float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write(
'Skip not unique mapping reads: %d(%.2f%s all reads)\n' %
(filter_not_mapping_reads,
float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique mapping reads:\n')
report_out.write('All unique mapping basepairs: %d\n' % all_mapping_bp)
report_out.write(
'Filter Duplicate reads: %d(%.2f%s of unique mapping reads)\n' %
(filter_duplicate_reads, float(filter_duplicate_reads) /
(all_reads - not_mapping_reads) * 100, "%"))
report_out.write(
"Filter 5' nonconversion basepairs: %d(%.2f%s of unique mapping basepairs)\n"
% (filter_nonuniform_trim_bp,
float(filter_nonuniform_trim_bp) / all_mapping_bp * 100, "%"))
report_out.write(
"Filter 5' nonconversion CpG basepairs: %d(%.2f%s of unique mapping basepairs)\n"
%
(filter_nonuniform_trim_bp_CG,
float(filter_nonuniform_trim_bp_CG) / all_mapping_bp * 100, "%"))
else:
report_out.write('Not unique paired mapping reads: %d(%.2f%s)\n' %
(not_mapping_reads,
float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write(
'Unique paired mapping reads: %d(%.2f%s)\n' %
((all_reads - not_mapping_reads),
float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write(
'Skip not paired unique mapping reads: %d(%.2f%s)\n' %
(filter_not_mapping_reads,
float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique paired mapping reads:\n')
report_out.write('All unique paired mapping basepairs: %d\n' %
all_mapping_bp)
report_out.write(
'Filter Duplicate reads: %d(%.2f%s of unique paired mapping reads)\n'
% (filter_duplicate_reads, float(filter_duplicate_reads) /
(all_reads - not_mapping_reads * 100), "%"))
report_out.write(
"Filter 5' nonconversion basepairs: %d(%.2f%s of unique mapping basepairs)\n"
% (filter_nonuniform_trim_bp,
float(filter_nonuniform_trim_bp) / all_mapping_bp * 100, "%"))
report_out.write(
"Filter 5' nonconversion CpG basepairs: %d(%.2f%s of unique mapping basepairs)\n"
%
(filter_nonuniform_trim_bp_CG,
float(filter_nonuniform_trim_bp_CG) / all_mapping_bp * 100, "%"))
report_out.write(
'Filter overlapped basepairs: %d(%.2f%s of unique paired mapping basepairs)\n'
% (filter_remove_overlap_bp,
float(filter_remove_overlap_bp) / all_mapping_bp * 100, "%"))
report_out.close()
info('Get the report file!')
def filter_sam(sam_inf, strand_t, read_l, single_on, name, s_path, auto, remove_overlap, loc_dict, max_cov,
not_mapping):
'''
Trim the mapping files with the biased positions of every length in every strand,
which are saved in the variance: strand_t.
'''
filter_duplicate_reads = 0
filter_mbias_trim_bp = 0
filter_remove_overlap_bp = 0
filter_not_mapping_reads = 0
all_reads = 0
not_mapping_reads = 0
all_mapping_bp = 0
for sam in sam_inf:
out_sam = sam[:-4] + '_' + name + '_filter.sam'
out = open(out_sam, 'w')
#check the input mapping files
sam_format, read_inf = check.check_mapping_file_header(sam, s_path)
#scan every read to qc_filter
for read in read_inf:
#for sam header
if read.startswith('@'):
out.write(read)
continue
else:
all_reads += 1 ##record the read number (2013-06-20)
#Get the read information for trimming
#If the read isn't unique mapping, we will get a empty list ([]).
#In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
#In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
read_info = RI(read)
read_info = read_info.extract_information()
if len(read_info) == 0:
not_mapping_reads += 1
if not_mapping: #keep the not_unique mapping reads
out.write(read)
else:
filter_not_mapping_reads += 1 ##record the not mapping read number (2013-06-20)
continue
if len(loc_dict) > 0: #the --filter_dup has been set True, have to remove duplicate reads
duplicate, loc_dict = DF(read_info, loc_dict, max_cov, single_on)
else:
duplicate = False
if single_on:
all_mapping_bp += len(read_info[5]) ##record the mapping read basepair (2013-06-20)
if auto:
if read_l[0] != '':
original_length = int(read_l[sam_inf.index(sam)])
else:
original_length = ''
filter_mbias_trim_bp, filter_duplicate_reads = SF(read, strand_t, out, read_info, original_length,
duplicate, filter_mbias_trim_bp, filter_duplicate_reads)
else:
if not duplicate and len(loc_dict) > 0:
out.write(read) #not trimming, only output not_duplicate reads
else:
filter_duplicate_reads += 1 ##record the duplicate read (2013-06-20)
else:
all_mapping_bp += len(read_info[7]) ##record the mapping read basepair (2013-06-20)
if auto or remove_overlap:
if read_l[0] != '':
original_length = [int(i) for i in read_l[sam_inf.index(sam)].split('-')]
else:
original_length = ''
filter_mbias_trim_bp, filter_duplicate_reads, filter_remove_overlap_bp = PF(read, strand_t, out,
read_info, original_length, auto, remove_overlap, duplicate, filter_mbias_trim_bp,
filter_duplicate_reads, filter_remove_overlap_bp)
else:
if not duplicate and len(loc_dict) > 0:
out.write(read) #not trimming, only output not_duplicate reads
else:
filter_duplicate_reads += 1 ##record the duplicate read (2013-06-20)
out.close()
##produce the filter report
info('Produce the report file...')
report_out = open(name + "_BSeQC_mbias_filter_report.txt", 'w')
report_out.write('Total reads: %d\n' % all_reads)
if single_on:
report_out.write('Not unique mapping reads: %d(%.2f%s all reads)\n' % (
not_mapping_reads, float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Unique mapping reads: %d(%.2f%s all reads)\n' % (
(all_reads - not_mapping_reads), float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Skip not unique mapping reads: %d(%.2f%s all reads)\n' % (
filter_not_mapping_reads, float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique mapping reads:\n')
report_out.write('All unique mapping basepairs: %d\n' % all_mapping_bp)
report_out.write('Filter Duplicate reads: %d(%.2f%s of unique mapping reads)\n' % (
filter_duplicate_reads, float(filter_duplicate_reads) / (all_reads - not_mapping_reads) * 100, "%"))
report_out.write('Filter Mbias basepairs: %d(%.2f%s of unique mapping basepairs)\n' % (
filter_mbias_trim_bp, float(filter_mbias_trim_bp) / all_mapping_bp * 100, "%"))
else:
report_out.write('Not unique paired mapping reads: %d(%.2f%s)\n' % (
not_mapping_reads, float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Unique paired mapping reads: %d(%.2f%s)\n' % (
(all_reads - not_mapping_reads), float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Skip not paired unique mapping reads: %d(%.2f%s)\n' % (
filter_not_mapping_reads, float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique paired mapping reads:\n')
report_out.write('All unique paired mapping basepairs: %d\n' % all_mapping_bp)
report_out.write('Filter Duplicate reads: %d(%.2f%s of unique paired mapping reads)\n' % (
filter_duplicate_reads, float(filter_duplicate_reads) / (all_reads - not_mapping_reads) * 100, "%"))
report_out.write('Filter Mbias basepairs: %d(%.2f%s of unique paired mapping basepairs)\n' % (
filter_mbias_trim_bp, float(filter_mbias_trim_bp) / all_mapping_bp * 100, "%"))
report_out.write('Filter overlapped basepairs: %d(%.2f%s of unique paired mapping basepairs)\n' % (
filter_remove_overlap_bp, float(filter_remove_overlap_bp) / all_mapping_bp * 100, "%"))
report_out.close()
info('Get the report file!')
def filter_sam(sam_inf, strand_t, read_l, single_on, name, s_path, auto, remove_overlap, loc_dict, max_cov,
not_mapping):
'''
Trim the mapping files with the biased positions of every length in every strand,
which are saved in the variance: strand_t.
'''
filter_duplicate_reads = 0
filter_mbias_trim_bp = 0
filter_remove_overlap_bp = 0
filter_not_mapping_reads = 0
all_reads = 0
not_mapping_reads = 0
all_mapping_bp = 0
for sam in sam_inf:
out_sam = sam[:-4] + '_' + name + '_filter.sam'
out = open(out_sam, 'w')
#check the input mapping files
sam_format, read_inf = check.check_mapping_file_header(sam, s_path)
#scan every read to qc_filter
for read in read_inf:
#for sam header
if read.startswith('@'):
out.write(read)
continue
else:
all_reads += 1 ##record the read number (2013-06-20)
#Get the read information for trimming
#If the read isn't unique mapping, we will get a empty list ([]).
#In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
#In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
read_info = RI(read)
read_info = read_info.extract_information()
if len(read_info) == 0:
not_mapping_reads += 1
if not_mapping: #keep the not_unique mapping reads
out.write(read)
else:
filter_not_mapping_reads += 1 ##record the not mapping read number (2013-06-20)
continue
if len(loc_dict) > 0: #the --filter_dup has been set True, have to remove duplicate reads
duplicate, loc_dict = DF(read_info, loc_dict, max_cov, single_on)
else:
duplicate = False
if single_on:
all_mapping_bp += len(read_info[5]) ##record the mapping read basepair (2013-06-20)
if auto:
if read_l[0] != '':
original_length = int(read_l[sam_inf.index(sam)])
else:
original_length = ''
filter_mbias_trim_bp, filter_duplicate_reads = SF(read, strand_t, out, read_info, original_length,
duplicate, filter_mbias_trim_bp, filter_duplicate_reads)
else:
if not duplicate and len(loc_dict) > 0:
out.write(read) #not trimming, only output not_duplicate reads
else:
filter_duplicate_reads += 1 ##record the duplicate read (2013-06-20)
else:
all_mapping_bp += len(read_info[7]) ##record the mapping read basepair (2013-06-20)
if auto or remove_overlap:
if read_l[0] != '':
original_length = [int(i) for i in read_l[sam_inf.index(sam)].split('-')]
else:
original_length = ''
filter_mbias_trim_bp, filter_duplicate_reads, filter_remove_overlap_bp = PF(read, strand_t, out,
read_info, original_length, auto, remove_overlap, duplicate, filter_mbias_trim_bp,
filter_duplicate_reads, filter_remove_overlap_bp)
else:
if not duplicate and len(loc_dict) > 0:
out.write(read) #not trimming, only output not_duplicate reads
else:
filter_duplicate_reads += 1 ##record the duplicate read (2013-06-20)
out.close()
##produce the filter report
info('Produce the report file...')
report_out = open(name + "_BSeQC_mbias_filter_report.txt", 'w')
report_out.write('Total reads: %d\n' % all_reads)
if single_on:
report_out.write('Not unique mapping reads: %d(%.2f%s all reads)\n' % (
not_mapping_reads, float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Unique mapping reads: %d(%.2f%s all reads)\n' % (
(all_reads - not_mapping_reads), float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Skip not unique mapping reads: %d(%.2f%s all reads)\n' % (
filter_not_mapping_reads, float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique mapping reads:\n')
report_out.write('All unique mapping basepairs: %d\n' % all_mapping_bp)
report_out.write('Filter Duplicate reads: %d(%.2f%s of unique mapping reads)\n' % (
filter_duplicate_reads, float(filter_duplicate_reads) / (all_reads - not_mapping_reads) * 100, "%"))
report_out.write('Filter Mbias basepairs: %d(%.2f%s of unique mapping basepairs)\n' % (
filter_mbias_trim_bp, float(filter_mbias_trim_bp) / all_mapping_bp * 100, "%"))
else:
report_out.write('Not unique paired mapping reads: %d(%.2f%s)\n' % (
not_mapping_reads, float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Unique paired mapping reads: %d(%.2f%s)\n' % (
(all_reads - not_mapping_reads), float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Skip not paired unique mapping reads: %d(%.2f%s)\n' % (
filter_not_mapping_reads, float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique paired mapping reads:\n')
report_out.write('All unique paired mapping basepairs: %d\n' % all_mapping_bp)
report_out.write('Filter Duplicate reads: %d(%.2f%s of unique paired mapping reads)\n' % (
filter_duplicate_reads, float(filter_duplicate_reads) / (all_reads - not_mapping_reads) * 100, "%"))
report_out.write('Filter Mbias basepairs: %d(%.2f%s of unique paired mapping basepairs)\n' % (
filter_mbias_trim_bp, float(filter_mbias_trim_bp) / all_mapping_bp * 100, "%"))
report_out.write('Filter overlapped basepairs: %d(%.2f%s of unique paired mapping basepairs)\n' % (
filter_remove_overlap_bp, float(filter_remove_overlap_bp) / all_mapping_bp * 100, "%"))
report_out.close()
info('Get the report file!')
def record_site(read, ref, bsm, dige_site):
'''
record the location of restriction enzyme digestion site in the read,
and the methylation level of the restriction enzyme digestion site
:param read:
:param ref:
:param dige_site:
'''
#Using the read class
#If the read is unique mapping or unique and paired mapping, we will get a information list.
#If not, we will get a empty list ([]).
#In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
#In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
read_s = RI(read, bsm)
read_info = read_s.extract_information()
strand = ''
site_meth_list = []
reverse_strand = ['-+', '+-']
if len(read_info) == 0:
return read_info, site_meth_list
strand, chr, pos, seq = read_info[1], read_info[2], int(read_info[3]) - 1, read_info[-2]
#print strand, chr, pos, seq
if len(chr.split('_')) > 1: #filter some 'chr_' chromosome
return read_info, site_meth_list
readlen = len(seq)
refseq = ref[chr][pos - 1: pos + readlen + 1]
if strand not in reverse_strand:
#get the location of restriction enzyme digestion site
dige_5_index = refseq.find('CCGG')
dige_3_index = refseq.rfind('CCGG')
else:
# if the strand is '-+' (single-end) or '+-', reverse the seq and refseq
seq = seq[::-1]
refseq = refseq[::-1]
dige_5_index = refseq.find('GGCC')
dige_3_index = refseq.rfind('GGCC')
if dige_3_index == -1 or dige_5_index == -1:
warning("Can't find MspI site in the read: %s" %read.rstrip())
#return dige_dict
return read_info, site_meth_list
# get the location of the restriction enzyme digestion site
#dige_dict[strand]['s'][0].append(dige_5_index + 1)
#dige_dict[strand]['e'][0].append(dige_3_index + 1)
# get the methylation state of the restriction enzyme digestion site
if strand[1] == '-':
#stand: +- or --, check the third nucleotide of C-CGG
dige_5_meth, dige_5_all = get_meth_state(strand, seq, refseq, dige_5_index + 1)
dige_3_meth, dige_3_all = get_meth_state(strand, seq, refseq, dige_3_index + 1)
else:
#stand: ++ or -+, check the second nucleotide of C-CGG
dige_5_meth, dige_5_all = get_meth_state(strand, seq, refseq, dige_5_index)
dige_3_meth, dige_3_all = get_meth_state(strand, seq, refseq, dige_3_index)
#dige_dict[strand]['s'][1] = [sum(x) for x in zip(dige_dict[strand]['s'][1], [dige_5_meth, dige_5_all])]
#dige_5_meth_list = [dige_5_meth] + [0] * 10
#dige_5_all_list = [dige_5_all] + [0] * 10
#dige_3_meth_list = [0] * 11
#dige_3_all_list = [0] * 11
#calculate the methylation states of 10 nucleotides after dige_5_index
#for i in range(1, 11):
#meth_5, all_5 = get_meth_state(strand, seq, refseq, dige_5_index + 2 + i) # +2 to scan the nucleotide after CCGG
#meth_3, all_3 = get_meth_state(strand, seq, refseq, len(seq) - i)
#dige_5_meth_list[i] = meth_5
#dige_5_all_list[i] = all_5
#dige_3_meth_list[10 - i] = meth_3
#dige_3_all_list[10 - i] = all_3
#site_meth_list = [dige_5_meth_list, dige_5_all_list, dige_3_meth_list, dige_3_all_list, dige_5_index, len(seq)]
site_meth_list = [dige_5_meth, dige_5_all, 0, 0, dige_5_index, len(seq)]
if dige_3_index != 0 and dige_3_index != dige_5_index:
if strand[1] == '+':
#check the 3' digestion site for '++' and '-+'
if len(seq[(dige_3_index - 1):(dige_3_index + 3)]) == 4:
if (strand == '++' and seq[dige_3_index + 2] == 'A') or (strand == '-+' and seq[dige_3_index + 2] == 'T'):
site_meth_list = [dige_5_meth, dige_5_all, dige_3_meth, dige_3_all, dige_5_index, dige_3_index]
#dige_dict[strand]['e'][1] = [sum(x) for x in zip(dige_dict[strand]['e'][1], [dige_3_meth, dige_3_all])]
#dige_dict[strand]['e'] = [sum(x) for x in zip(dige_dict[strand]['e'], [dige_3_meth, dige_3_all])]
#for i in range(1, 11):
# meth_3, all_3 = get_meth_state(strand, seq, refseq, dige_3_index - 1 - i)
# dige_3_meth_list[10 - i] = meth_3
# dige_3_all_list[10 - i] = all_3
#dige_3_meth_list[-1] = dige_3_meth
#dige_3_all_list[-1] = dige_3_all
#site_meth_list = [dige_5_meth_list, dige_5_all_list, dige_3_meth_list, dige_3_all_list, dige_5_index, dige_3_index]
#print "end_repair_3", refseq[dige_3_index:(dige_3_index + 4)], dige_3_meth_list, dige_3_all_list
else:
site_meth_list = [dige_5_meth, dige_5_all, dige_3_meth, dige_3_all, dige_5_index, dige_3_index]
#dige_dict[strand]['e'][1] = [sum(x) for x in zip(dige_dict[strand]['e'][1], [dige_3_meth, dige_3_all])]
#dige_dict[strand]['e'] = [sum(x) for x in zip(dige_dict[strand]['e'], [dige_3_meth, dige_3_all])]
#for i in range(1, 11):
# meth_3, all_3 = get_meth_state(strand, seq,refseq, dige_3_index - 1 - i)
# dige_3_meth_list[10 - i] = meth_3
# dige_3_all_list[10 - i] = all_3
#dige_3_meth_list[-1] = dige_3_meth
#dige_3_all_list[-1] = dige_3_all
#site_meth_list = [dige_5_meth_list, dige_5_all_list, dige_3_meth_list, dige_3_all_list, dige_5_index, dige_3_index]
#print 'bbbb', dige_3_meth_list, dige_3_all_list
#return dige_dict
return read_info, site_meth_list
def record_site(read, ref, bsm, dige_site):
'''
record the location of restriction enzyme digestion site in the read,
and the methylation level of the restriction enzyme digestion site
:param read:
:param ref:
:param dige_site:
'''
#Using the read class
#If the read is unique mapping or unique and paired mapping, we will get a information list.
#If not, we will get a empty list ([]).
#In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
#In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
read_s = RI(read, bsm)
read_info = read_s.extract_information()
strand = ''
site_meth_list = []
reverse_strand = ['-+', '+-']
if len(read_info) == 0:
return read_info, site_meth_list
strand, chr, pos, seq = read_info[1], read_info[2], int(
read_info[3]) - 1, read_info[-2]
#print strand, chr, pos, seq
if len(chr.split('_')) > 1: #filter some 'chr_' chromosome
return read_info, site_meth_list
readlen = len(seq)
refseq = ref[chr][pos - 1:pos + readlen + 1]
if strand not in reverse_strand:
#get the location of restriction enzyme digestion site
dige_5_index = refseq.find('CCGG')
dige_3_index = refseq.rfind('CCGG')
else:
# if the strand is '-+' (single-end) or '+-', reverse the seq and refseq
seq = seq[::-1]
refseq = refseq[::-1]
dige_5_index = refseq.find('GGCC')
dige_3_index = refseq.rfind('GGCC')
if dige_3_index == -1 or dige_5_index == -1:
warning("Can't find MspI site in the read: %s" % read.rstrip())
#return dige_dict
return read_info, site_meth_list
# get the location of the restriction enzyme digestion site
#dige_dict[strand]['s'][0].append(dige_5_index + 1)
#dige_dict[strand]['e'][0].append(dige_3_index + 1)
# get the methylation state of the restriction enzyme digestion site
if strand[1] == '-':
#stand: +- or --, check the third nucleotide of C-CGG
dige_5_meth, dige_5_all = get_meth_state(strand, seq, refseq,
dige_5_index + 1)
dige_3_meth, dige_3_all = get_meth_state(strand, seq, refseq,
dige_3_index + 1)
else:
#stand: ++ or -+, check the second nucleotide of C-CGG
dige_5_meth, dige_5_all = get_meth_state(strand, seq, refseq,
dige_5_index)
dige_3_meth, dige_3_all = get_meth_state(strand, seq, refseq,
dige_3_index)
#dige_dict[strand]['s'][1] = [sum(x) for x in zip(dige_dict[strand]['s'][1], [dige_5_meth, dige_5_all])]
#dige_5_meth_list = [dige_5_meth] + [0] * 10
#dige_5_all_list = [dige_5_all] + [0] * 10
#dige_3_meth_list = [0] * 11
#dige_3_all_list = [0] * 11
#calculate the methylation states of 10 nucleotides after dige_5_index
#for i in range(1, 11):
#meth_5, all_5 = get_meth_state(strand, seq, refseq, dige_5_index + 2 + i) # +2 to scan the nucleotide after CCGG
#meth_3, all_3 = get_meth_state(strand, seq, refseq, len(seq) - i)
#dige_5_meth_list[i] = meth_5
#dige_5_all_list[i] = all_5
#dige_3_meth_list[10 - i] = meth_3
#dige_3_all_list[10 - i] = all_3
#site_meth_list = [dige_5_meth_list, dige_5_all_list, dige_3_meth_list, dige_3_all_list, dige_5_index, len(seq)]
site_meth_list = [dige_5_meth, dige_5_all, 0, 0, dige_5_index, len(seq)]
if dige_3_index != 0 and dige_3_index != dige_5_index:
if strand[1] == '+':
#check the 3' digestion site for '++' and '-+'
if len(seq[(dige_3_index - 1):(dige_3_index + 3)]) == 4:
if (strand == '++' and seq[dige_3_index + 2] == 'A') or (
strand == '-+' and seq[dige_3_index + 2] == 'T'):
site_meth_list = [
dige_5_meth, dige_5_all, dige_3_meth, dige_3_all,
dige_5_index, dige_3_index
]
#dige_dict[strand]['e'][1] = [sum(x) for x in zip(dige_dict[strand]['e'][1], [dige_3_meth, dige_3_all])]
#dige_dict[strand]['e'] = [sum(x) for x in zip(dige_dict[strand]['e'], [dige_3_meth, dige_3_all])]
#for i in range(1, 11):
# meth_3, all_3 = get_meth_state(strand, seq, refseq, dige_3_index - 1 - i)
# dige_3_meth_list[10 - i] = meth_3
# dige_3_all_list[10 - i] = all_3
#dige_3_meth_list[-1] = dige_3_meth
#dige_3_all_list[-1] = dige_3_all
#site_meth_list = [dige_5_meth_list, dige_5_all_list, dige_3_meth_list, dige_3_all_list, dige_5_index, dige_3_index]
#print "end_repair_3", refseq[dige_3_index:(dige_3_index + 4)], dige_3_meth_list, dige_3_all_list
else:
site_meth_list = [
dige_5_meth, dige_5_all, dige_3_meth, dige_3_all, dige_5_index,
dige_3_index
]
#dige_dict[strand]['e'][1] = [sum(x) for x in zip(dige_dict[strand]['e'][1], [dige_3_meth, dige_3_all])]
#dige_dict[strand]['e'] = [sum(x) for x in zip(dige_dict[strand]['e'], [dige_3_meth, dige_3_all])]
#for i in range(1, 11):
# meth_3, all_3 = get_meth_state(strand, seq,refseq, dige_3_index - 1 - i)
# dige_3_meth_list[10 - i] = meth_3
# dige_3_all_list[10 - i] = all_3
#dige_3_meth_list[-1] = dige_3_meth
#dige_3_all_list[-1] = dige_3_all
#site_meth_list = [dige_5_meth_list, dige_5_all_list, dige_3_meth_list, dige_3_all_list, dige_5_index, dige_3_index]
#print 'bbbb', dige_3_meth_list, dige_3_all_list
#return dige_dict
return read_info, site_meth_list
def run(args):
"""
Alternative module: Use the strategy in Bis-SNP to trim 5' bisulfite conversion failures
"""
options = args.parse_args()
if len(options.sam_file) == 0:
error("Missing the SAM file, use -s or --sam option.")
else:
options.sam_file = options.sam_file.split(',')
for s in options.sam_file:
if not os.path.isfile(s):
error("Can't open the SAM file: " + s)
sys.exit(1)
if len(options.ref_file) == 0:
error("Missing the reference genome fasta file, use -r or --ref option.")
else:
if not os.path.isfile(options.ref_file):
error("Can't open the ref file: " + options.ref_file)
if len(options.samtools) != 0:
if options.samtools[-1] != '/':
options.samtools += '/'
if len(options.name) == 0:
error("Missing the output file name, use -n or --name options.")
sam_inf = options.sam_file
ref_file = options.ref_file
bsm = options.bsm
s_path = options.samtools
name = options.name
remove_overlap = options.remove_overlap
filter_dup = options.filter_dup
p_poisson = options.p_poisson
gsize = options.gsize
not_mapping = options.not_mapping
info("Get the all parameter!!")
#check the input mapping files
sam_format, read_inf = check.check_mapping_file_flag(sam_inf[0], s_path)
pre_flag = read_inf.readline().split('\t')[1]
if 'p' in pre_flag:
single_on = False
info("The input mapping files are paired-end sequencing!")
else:
single_on = True
info("The input mapping files are single-end sequencing!")
loc_dict = {}
if filter_dup:
## if filter_up is TRUE, the duplicate reads will be assessed and shown in Dup_dis.pdf
info("The filter_dup has been set True.")
info("Assess the duplicate reads...")
for sam in sam_inf:
#check the input mapping files
sam_format, read_inf = check.check_mapping_file(sam, s_path)
if single_on:
for read in read_inf:
loc_dict = LI.Loc_single(read, loc_dict, bsm)
else:
for read in read_inf:
loc_dict = LI.Loc_paired(read, loc_dict, bsm)
max_cov = DR.duplicate_report(loc_dict, gsize, p_poisson, name)
info('Get the duplicate reads distribution!')
#get reference information
ref = GR.get_ref(ref_file)
trim_position = []
filter_duplicate_reads = 0
filter_nonuniform_trim_bp = 0
filter_nonuniform_trim_bp_CG = 0
filter_remove_overlap_bp = 0
filter_not_mapping_reads = 0
all_reads = 0
not_mapping_reads = 0
all_mapping_bp = 0
##filter the 5' bisulfite failure
for sam in sam_inf:
out_sam = sam[:-4] + '_' + name + '_filter.sam'
out = open(out_sam, 'w')
#check the input mapping files
record_mate = {}
sam_format, read_inf = check.check_mapping_file_header(sam, s_path)
for read in read_inf:
#for sam header
if read.startswith('@'):
out.write(read)
continue
else:
all_reads += 1 ##record the read number (2013-06-20)
#Get the read information for trimming
#If the read isn't unique mapping, we will get a empty list ([]).
#In: single unique mapping read Out: [flag,strand,chr,pos,CIGAR,seq,score]
#In: paired unique mapping read Out: [flag,strand,chr,pos1,CIGAR,pos2,insert,seq,score]
read_info = RI(read, bsm)
read_info = read_info.extract_information()
if len(read_info) == 0:
not_mapping_reads += 1
if not_mapping: #keep the not_unique mapping reads (or not paired mapping)
out.write(read)
else:
filter_not_mapping_reads += 1 ##record the not mapping read number (2013-06-20)
continue
if len(loc_dict) > 0: #the --filter_dup has been set True, have to remove duplicate reads
duplicate, loc_dict = DF(read_info, loc_dict, max_cov, single_on)
else:
duplicate = False
if single_on:
all_mapping_bp += len(read_info[5]) ##record the mapping read basepair (2013-06-20)
else:
all_mapping_bp += len(read_info[7]) ##record the mapping read basepair (2013-06-20)
record_mate, trim_position, filter_nonuniform_trim_bp_CG, filter_duplicate_reads, filter_remove_overlap_bp = NF.nonuniform_filter(read,
out,
read_info,
ref,
remove_overlap,
duplicate,
single_on,
record_mate,
trim_position,
filter_nonuniform_trim_bp_CG,
filter_duplicate_reads,
filter_remove_overlap_bp)
out.close()
del record_mate
NR.nonuniform_generator(trim_position, name)
for i in range(len(trim_position)):
filter_nonuniform_trim_bp += i * trim_position[i]
##produce the filter report
info('Produce the report file...')
report_out = open(name + "_BSeQC_nonuniform_filter_report.txt", 'w')
report_out.write('Total reads: %d\n' % all_reads)
if single_on:
report_out.write('Not unique mapping reads: %d(%.2f%s all reads)\n' % (
not_mapping_reads, float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Unique mapping reads: %d(%.2f%s all reads)\n' % (
(all_reads - not_mapping_reads), float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Skip not unique mapping reads: %d(%.2f%s all reads)\n' % (
filter_not_mapping_reads, float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique mapping reads:\n')
report_out.write('All unique mapping basepairs: %d\n' % all_mapping_bp)
report_out.write('Filter Duplicate reads: %d(%.2f%s of unique mapping reads)\n' % (
filter_duplicate_reads, float(filter_duplicate_reads) / (all_reads - not_mapping_reads) * 100, "%"))
report_out.write("Filter 5' nonconversion basepairs: %d(%.2f%s of unique mapping basepairs)\n" % (
filter_nonuniform_trim_bp, float(filter_nonuniform_trim_bp) / all_mapping_bp * 100, "%"))
report_out.write("Filter 5' nonconversion CpG basepairs: %d(%.2f%s of unique mapping basepairs)\n" % (
filter_nonuniform_trim_bp_CG, float(filter_nonuniform_trim_bp_CG) / all_mapping_bp * 100, "%"))
else:
report_out.write('Not unique paired mapping reads: %d(%.2f%s)\n' % (
not_mapping_reads, float(not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Unique paired mapping reads: %d(%.2f%s)\n' % (
(all_reads - not_mapping_reads), float(all_reads - not_mapping_reads) / all_reads * 100, "%"))
report_out.write('Skip not paired unique mapping reads: %d(%.2f%s)\n' % (
filter_not_mapping_reads, float(filter_not_mapping_reads) / all_reads * 100, "%"))
report_out.write('In unique paired mapping reads:\n')
report_out.write('All unique paired mapping basepairs: %d\n' % all_mapping_bp)
report_out.write('Filter Duplicate reads: %d(%.2f%s of unique paired mapping reads)\n' % (
filter_duplicate_reads, float(filter_duplicate_reads) / (all_reads - not_mapping_reads * 100), "%"))
report_out.write("Filter 5' nonconversion basepairs: %d(%.2f%s of unique mapping basepairs)\n" % (
filter_nonuniform_trim_bp, float(filter_nonuniform_trim_bp) / all_mapping_bp * 100, "%"))
report_out.write("Filter 5' nonconversion CpG basepairs: %d(%.2f%s of unique mapping basepairs)\n" % (
filter_nonuniform_trim_bp_CG, float(filter_nonuniform_trim_bp_CG) / all_mapping_bp * 100, "%"))
report_out.write('Filter overlapped basepairs: %d(%.2f%s of unique paired mapping basepairs)\n' % (
filter_remove_overlap_bp, float(filter_remove_overlap_bp) / all_mapping_bp * 100, "%"))
report_out.close()
info('Get the report file!')
