def main(argv=None):
settings = process_command_line(argv)
try:
pos_feat_list, all_features = RILseq.read_gtf(
open(settings.genes_gff), settings.feature, settings.identifier)
except IOError:
return 1
lib_order = []
all_counts = {}
if settings.singles:
settings.only_first = True
settings.only_second = False
for r1_name in RILseq.flat_list(settings.reads_files):
sys.stderr.write('%s\n'%str(r1_name))
lib_order.append(r1_name)
all_counts[r1_name] = count_features(
pos_feat_list, open(r1_name), settings.overlap, length=25,
ignore_first=settings.only_second,
ignore_second=settings.only_first, count_singles=settings.singles)
outt = csv.writer(sys.stdout, delimiter='\t')
if not settings.quiet:
outt.writerow(['Gene name'] + lib_order)
for g in sorted(list(all_features)):
if settings.quiet and g.startswith('~'):
continue
row_out = [g]
for libn in lib_order:
row_out.append(all_counts[libn][g])
outt.writerow(row_out)
# application code here, like:
# run(settings, args)
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
if not os.path.exists(settings.dirout):
os.makedirs(settings.dirout)
if settings.genes_gff:
try:
pos_feat_list, all_features = RILseq.read_gtf(
open(settings.genes_gff), settings.feature, settings.identifier)
except IOError:
settings.genes_gff = None
gcounts = {}
lib_order = []
fastq_1_list = list(RILseq.flat_list(settings.fastq_1))
fastq_2_list = list(RILseq.flat_list(settings.fastq_2))
for i, r1_name in enumerate(RILseq.flat_list(settings.fastq_1)):
try:
r2_name = fastq_2_list[i]
except IndexError:
r2_name = None
outhead = r1_name.rsplit('.', 1)[0]
libname = outhead.rsplit('/',1)[-1]
outhead = '%s_bwa'%libname
bamname = RILseq.run_bwa(
settings.bwa_exec, r1_name, r2_name,
settings.dirout, outhead, settings.allowed_mismatches,
settings.genome_fasta, settings.params_aln, settings.sampe_params,
settings.samse_params, settings.samtools_cmd)
samfile = pysam.Samfile(bamname)
if settings.genes_gff:
lib_order.append(libname)
gcounts[libname] = RILseq.count_features(
pos_feat_list, samfile, settings.overlap,
rev=settings.reverse_complement)
if settings.create_wig:
outwigs = [open("%s/%s_coverage.wig"%(settings.dirout, fastq.split("_cutadapt")[0]), 'w')
for fastq in fastq_1_list]
coverage = RILseq.generate_wig(
samfile, rev=settings.reverse_complement, first_pos=False)
RILseq.print_wiggle(
coverage, "%s_single_fragments_coverage"%libname,
"%s single fragments coverage"%libname, outwigs[i])
# Print the table of counts
if settings.genes_gff:
outtables = [open("%s/%s_counts.txt"%(settings.dirout, fastq.split("_cutadapt")[0]), 'w')
for fastq in fastq_1_list]
for i, r1_name in enumerate(fastq_1_list):
outt = csv.writer(outtables[i], delimiter='\t')
outt.writerow(['Gene name'] + lib_order)
for g in sorted(list(all_features)):
row_out = [g]
for libn in lib_order:
row_out.append(gcounts[libn][g])
outt.writerow(row_out)
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
# Read the transcripts if given
if settings.transcripts:
trans_dict = RILseq.read_transcripts(settings.transcripts)
else:
trans_dict = None
# Get the ends of the reads from the bam files
# sys.stderr.write('%s\n'%str(settings.bamfiles))
if settings.all_reads:
try:
outall = open(settings.all_reads, 'w')
except IOError:
outall = None
elif settings.add_all_reads:
outall = sys.stdout
else:
outall = None
for bf in RILseq.flat_list(settings.bamfiles):
bfin = pysam.Samfile(bf)
outhead = bf.rsplit('.', 1)[0]
libname = outhead.rsplit('/',1)[-1]
fsq1name = "%s/%s_ends_1.fastq"%(settings.dirout, libname)
fsq2name = "%s/%s_ends_2.fastq"%(settings.dirout, libname)
if settings.skip_mapping:
fsq1 = open(os.devnull, 'w')
fsq2 = fsq1
else:
fsq1 = open(fsq1name, 'w')
fsq2 = open(fsq2name, 'w')
single_mapped = RILseq.get_unmapped_reads(
bfin, fsq1, fsq2, settings.length, settings.maxG,
rev=settings.reverse_complement, all_reads=True,
dust_thr=settings.dust_thr)
reads_in = []
# Map the fastq files to the genome
for fqname in (fsq1name, fsq2name):
bamheadname = fqname.rsplit('.',1)[0].rsplit('/',1)[-1]
if settings.skip_mapping:
bamname = "%s/%s.bam"%(settings.dirout, bamheadname)
else:
bamname = RILseq.run_bwa(
settings.bwa_exec, fqname, None,
settings.dirout, bamheadname, settings.max_mismatches,
settings.genome_fasta, settings.params_aln,
'', settings.samse_params,
settings.samtools_cmd, processors=settings.processors)
bamin = pysam.Samfile(bamname)
reads_in.append(RILseq.read_bam_file(
bamin, bamin.references, settings.allowed_mismatches))
RILseq.write_reads_table(
sys.stdout, reads_in[0], reads_in[1], bfin.references,
settings.distance, not settings.keep_circular,
trans_dict, write_single=outall, single_mapped=single_mapped,
max_NM=settings.allowed_mismatches)
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
if not os.path.exists(settings.dirout):
os.makedirs(settings.dirout)
outwig = open("%s/%s_coverage.wig"%(settings.dirout, settings.outhead), 'w')
if settings.genes_gff:
try:
pos_feat_list, all_features = RILseq.read_gtf(
open(settings.genes_gff), settings.feature, settings.identifier)
except IOError:
settings.genes_gff = None
gcounts = {}
lib_order = []
fastq_2_list = list(RILseq.flat_list(settings.fastq_2))
for i, r1_name in enumerate(RILseq.flat_list(settings.fastq_1)):
try:
r2_name = fastq_2_list[i]
except IndexError:
r2_name = None
outhead = r1_name.rsplit('.', 1)[0]
libname = outhead.rsplit('/',1)[-1]
outhead = '%s_bwa'%libname
bamname = RILseq.run_bwa(
settings.bwa_exec, r1_name, r2_name,
settings.dirout, outhead, settings.allowed_mismatches,
settings.genome_fasta, settings.params_aln, settings.sampe_params,
settings.samse_params, settings.samtools_cmd,
processors=settings.processors)
samfile = pysam.Samfile(bamname)
if settings.genes_gff:
lib_order.append(libname)
gcounts[libname] = RILseq.count_features(
pos_feat_list, samfile, settings.overlap,
rev=settings.reverse_complement)
coverage = RILseq.generate_wig(
samfile, rev=settings.reverse_complement, first_pos=False)
RILseq.print_wiggle(
coverage, "%s_single_fragments_coverage"%libname,
"%s single fragments coverage"%libname, outwig)
# Print the table of counts
if settings.genes_gff:
outtable = open(
"%s/%s_counts.txt"%(settings.dirout, settings.outhead), 'w')
outt = csv.writer(outtable, delimiter='\t')
outt.writerow(['Gene name'] + lib_order)
for g in sorted(list(all_features)):
row_out = [g]
for libn in lib_order:
row_out.append(gcounts[libn][g])
outt.writerow(row_out)
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
try:
pos_feat_list, all_features = RILseq.read_gtf(open(settings.genes_gff),
settings.feature,
settings.identifier)
except IOError:
return 1
lib_order = []
all_counts = {}
if settings.singles:
settings.only_first = True
settings.only_second = False
for r1_name in RILseq.flat_list(settings.reads_files):
sys.stderr.write('%s\n' % str(r1_name))
lib_order.append(r1_name)
all_counts[r1_name] = count_features(pos_feat_list,
open(r1_name),
settings.overlap,
length=25,
ignore_first=settings.only_second,
ignore_second=settings.only_first,
count_singles=settings.singles)
outt = csv.writer(sys.stdout, delimiter='\t')
if not settings.quiet:
outt.writerow(['Gene name'] + lib_order)
for g in sorted(list(all_features)):
if settings.quiet and g.startswith('~'):
continue
row_out = [g]
for libn in lib_order:
row_out.append(all_counts[libn][g])
outt.writerow(row_out)
# application code here, like:
# run(settings, args)
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
if settings.ribozero and settings.bc_dir:
try:
# sys.stderr.write("{}\n".format(','.join(settings.rrna_list)))
uid_pos,_,_,_,_,rRNAs = RILseq.ecocyc_parser.read_genes_data(
settings.bc_dir, rRNA_prod=settings.rrna_list)
except IOError:
raise
rr_pos = []
chr_dict = dict(zip(
settings.BC_chrlist.split(',')[1::2],
settings.BC_chrlist.split(',')[0::2]))
for rrgene in rRNAs:
# Pad the position of the rRNA gene with the alignment length
rr_pos.append([chr_dict[uid_pos[rrgene][0]]] +\
[uid_pos[rrgene][1]-settings.length] +\
[uid_pos[rrgene][2]+settings.length] +\
[uid_pos[rrgene][3]])
# sys.stderr.write('\n'.join(rRNAs))
# print uid_pos.items()[:10]
# sys.stderr.write('\n'.join(' '.join(map(str,sl)) for sl in rr_pos))
else:
rr_pos = None
region_interactions, region_ints_as1, region_ints_as2, total_interactions=\
RILseq.read_reads_table(
open(settings.reads_in), settings.seglen, rr_pos, settings.only_singles)
# If all interactions are desired, skip the tests and report all
if settings.all_interactions:
interacting_regions = []
for reg1, r1data in region_interactions.items():
for reg2, clist in r1data.items():
interacting_regions.append(
(1, len(clist), 0, reg1[0], reg1[0]+settings.seglen,
reg1[1], reg1[2], reg2[0], reg2[0]+settings.seglen,
reg2[1], reg2[2], 0, 0, 0))
else:
# Now run the test for each pair of interacting regions
found_in_interaction = defaultdict(bool)
interacting_regions = []
# Start with the regions with the most interactions
pairs_num = {}
for reg1 in list(region_interactions.keys()):
if region_ints_as1[reg1] < settings.min_int:
continue
for reg2 in list(region_interactions[reg1].keys()):
if len(region_interactions[reg1][reg2]) >= settings.min_int:
pairs_num[(reg1, reg2)] = len(region_interactions[reg1][reg2])
# Iterate the list of regions from the pairs with many interactions down
for (reg1, reg2) in sorted(pairs_num, key=pairs_num.get, reverse=True):
pv, ints, odds, r1_from, r1_to, r2_from, r2_to, mat_b, mat_c,mat_d=\
RILseq.minpv_regions(
reg1, reg2, region_interactions, region_ints_as1,
region_ints_as2, total_interactions, found_in_interaction,
settings.seglen, settings.maxsegs, settings.min_odds_ratio)
pv *= len(pairs_num)
if pv <= settings.max_pv:
# Mark as participating
for r1 in range(r1_from, r1_to, settings.seglen):
for r2 in range(r2_from, r2_to, settings.seglen):
found_in_interaction[
(r1, reg1[1], reg1[2], r2, reg2[1], reg2[2])] = True
# Report this interaction
interacting_regions.append(
(pv, ints, odds, r1_from, r1_to, reg1[1], reg1[2], r2_from,
r2_to, reg2[1], reg2[2], mat_b, mat_c, mat_d))
# Read the number of total RNAs in each region if the bam file is given
sum_reads=0
if settings.total_RNA:
# prepare a dictionary of features
feat_dict = defaultdict(lambda: defaultdict(set))
for region1 in region_ints_as1:
for i in range(settings.seglen):
feat_dict[region1[2]+region1[1]][region1[0]+i].add(region1)
for region2 in region_ints_as2:
for i in range(settings.seglen):
feat_dict[region2[2]+region2[1]][region2[0]+i].add(region2)
feat_list = {}
for chrom, data in feat_dict.items():
maxpos = max(data.keys())
list_of_sets = []
for k in range(maxpos+1):
list_of_sets.append(list(data[k]))
feat_list[chrom] = list_of_sets
totRNA_counts = defaultdict(int)
sum_reads = 0
for bamfile in settings.total_RNA.split(','):
saminf = pysam.Samfile(bamfile)
totcounts, sum_of_counts_lib = RILseq.count_features(
feat_list, saminf, 5,
rev=settings.total_reverse, get_sum=True)
for k, v in totcounts.items():
totRNA_counts[k] += v
sum_reads += sum_of_counts_lib
# Collect all the ratios between IP and total then choose the 90%
# percentile to avoid liers
max_IP_div_total_as1 = []
max_IP_div_total_as2 = []
for reg, counts in totRNA_counts.items():
if counts > settings.min_total_counts:
counts = float(counts+1)
# div_prod = max(region_ints_as1[reg]/counts, region_ints_as2[reg]/counts)
max_IP_div_total_as1.append(region_ints_as1[reg]/counts)
max_IP_div_total_as2.append(region_ints_as2[reg]/counts)
# mm1_sorted = sorted(max_IP_div_total_as1)
mm_sorted = sorted(max_IP_div_total_as2+max_IP_div_total_as1)
max_IP_div_total = mm_sorted[
int(len(mm_sorted)*settings.norm_percentile)]
sys.stderr.write("%f\n"%(max_IP_div_total))
else:
totRNA_counts = defaultdict(int)
max_IP_div_total = 0
if (settings.shuffles ==0 and settings.run_RNAup):
settings.shuffles=-1
# Read the additional data to decorate the results with
RILseq.report_interactions(
region_interactions, sys.stdout, interacting_regions, settings.seglen,
settings.bc_dir, settings.genome, settings.BC_chrlist, settings.refseq_dir,
settings.targets_file, settings.rep_table, settings.single_counts,
settings.shuffles, settings.RNAup_cmd, settings.servers,
settings.length, settings.est_utr_lens, settings.pad_seqs,
totRNA_counts, max_IP_div_total, total_interactions, sum_reads, settings.linear_chromosome_list)
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
# Read the read names and positions
read_5ps = {}
read_3ps = {}
read_genes = {}
genome = {}
gsize = {}
for sr in SeqIO.parse(settings.genome, 'fasta'):
genome[sr.id] = sr.seq
gsize[sr.id] = len(sr.seq)
if len(settings.EC_chrlist) >= 2:
chr_dict = dict(zip(
settings.EC_chrlist.split(',')[0::2],
settings.EC_chrlist.split(',')[1::2]))
else:
chr_dict = {}
if settings.summary:
sig_reads = RILseq.read_significant_reads(
settings.summary, chr_dict, gname=settings.gene_name)
for line in csv.reader(open(settings.list_reads), delimiter='\t'):
# skip single
if len(line) > 7 and line[7]=="single":
continue
if settings.summary:
if (int(line[4])-1, line[5], line[3]) not in\
sig_reads[(int(line[1])-1, line[2], line[0])]:
continue
read_5ps[line[6]] = [int(line[1])-1, line[2], line[0]]
read_3ps[line[6]] = [int(line[4])-1, line[5], line[3]]
# read_genes[line[6]] = [line[0], line[1]]
# Read the bam files and return the long sequences
r1_seqs = {}
r2_seqs = {}
for bamfile in list(RILseq.flat_list(settings.bamfiles)):
r1s, r2s = get_reads_seqs(
pysam.Samfile(bamfile), read_5ps.keys(), rev=settings.reverse)
r1_seqs.update(r1s)
r2_seqs.update(r2s)
# For each read find the overlap, if exists and find the fusion point
outer = csv.writer(sys.stdout, delimiter='\t')
print 'track name="%s" description="%s" visibility=4 itemRgb="On" useScore=0'%(
settings.track_name, settings.track_desc)
# Because I'm lazy, the code is written so r1 is the 3' end of the fragment
for rname in set(r2_seqs.keys()):
if rname in r1_seqs:
r2seq = r2_seqs[rname]
r1seq = r1_seqs[rname]
else: # single-end
r2seq = r2_seqs[rname]
r1seq = ''
s1, overlap, s2 = find_overlap(r2seq, r1seq)
side_5p_len = extend_alignment(
s1+overlap+s2, read_5ps[rname][0], 0, False, read_5ps[rname][1],
genome[read_5ps[rname][2]])
side_3p_len = extend_alignment(
s1+overlap+s2, 0, read_3ps[rname][0], True, read_3ps[rname][1],
genome[read_3ps[rname][2]])
# Write each of the sides to the output file
score=0
if settings.rand_score:
score=random.randint(0, 1000)
if read_5ps[rname][1] == '+':
gfrom = max(0, read_5ps[rname][0])
gto = min(gsize[read_5ps[rname][2]], read_5ps[rname][0]+side_5p_len)
outer.writerow([
read_5ps[rname][2], gfrom, gto, "%s_5p"%rname, score, '+',
gfrom, gto, settings.pos_first])
elif read_5ps[rname][1] == '-':
gfrom = max(0, read_5ps[rname][0]-side_5p_len+1)
gto = min(gsize[read_5ps[rname][2]], read_5ps[rname][0]+1)
outer.writerow([
read_5ps[rname][2], gfrom, gto, "%s_5p"%rname, score, '-',
gfrom, gto,settings.rev_first])
if read_3ps[rname][1] == '+':
gfrom = max(0, read_3ps[rname][0]-side_3p_len+1)
gto = min(gsize[read_3ps[rname][2]], read_3ps[rname][0]+1)
outer.writerow([
read_3ps[rname][2], gfrom, gto,"%s_3p"%rname, score, '+',
gfrom, gto, settings.pos_second])
elif read_3ps[rname][1] == '-':
gfrom = max(0, read_3ps[rname][0])
gto = min(gsize[read_3ps[rname][2]], read_3ps[rname][0]+side_3p_len)
outer.writerow([
read_3ps[rname][2], gfrom, gto, "%s_3p"%rname, score, '-',
gfrom, gto, settings.rev_second])
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
if settings.ribozero:
try:
uid_pos,_,_,_,_,rRNAs = RILseq.ecocyc_parser.read_genes_data(
settings.ec_dir)
except IOError:
raise
rr_pos = []
chr_dict = dict(zip(
settings.EC_chrlist.split(',')[1::2],
settings.EC_chrlist.split(',')[0::2]))
for rrgene in rRNAs:
# Pad the position of the rRNA gene with the alignment lenght
rr_pos.append([chr_dict[uid_pos[rrgene][0]]] +\
[uid_pos[rrgene][1]-settings.length] +\
[uid_pos[rrgene][2]+settings.length] +\
[uid_pos[rrgene][3]])
else:
rr_pos = None
region_interactions, region_ints_as1, region_ints_as2, total_interactions=\
RILseq.read_reads_table(
open(settings.reads_in), settings.seglen, rr_pos, settings.only_singles)
# If all interactions are desired, skip the tests and report all
if settings.all_interactions:
interacting_regions = []
for reg1, r1data in region_interactions.items():
for reg2, clist in r1data.items():
interacting_regions.append(
(1, len(clist), 0, reg1[0], reg1[0]+settings.seglen,
reg1[1], reg1[2], reg2[0], reg2[0]+settings.seglen,
reg2[1], reg2[2], 0, 0, 0))
else:
# Now run the test for each pair of interacting regions
found_in_interaction = defaultdict(bool)
interacting_regions = []
# Start with the regions with the most interactions
pairs_num = {}
for reg1 in list(region_interactions.keys()):
if region_ints_as1[reg1] < settings.min_int:
continue
for reg2 in list(region_interactions[reg1].keys()):
if len(region_interactions[reg1][reg2]) >= settings.min_int:
pairs_num[(reg1, reg2)] = len(region_interactions[reg1][reg2])
# Iterate the list of regions from the pairs with many interactions down
for (reg1, reg2) in sorted(pairs_num, key=pairs_num.get, reverse=True):
pv, ints, odds, r1_from, r1_to, r2_from, r2_to, mat_b, mat_c,mat_d=\
RILseq.minpv_regions(
reg1, reg2, region_interactions, region_ints_as1,
region_ints_as2, total_interactions, found_in_interaction,
settings.seglen, settings.maxsegs, settings.min_odds_ratio)
pv *= len(pairs_num)
if pv <= settings.max_pv:
# Mark as participating
for r1 in range(r1_from, r1_to, settings.seglen):
for r2 in range(r2_from, r2_to, settings.seglen):
found_in_interaction[
(r1, reg1[1], reg1[2], r2, reg2[1], reg2[2])] = True
# Report this interaction
interacting_regions.append(
(pv, ints, odds, r1_from, r1_to, reg1[1], reg1[2], r2_from,
r2_to, reg2[1], reg2[2], mat_b, mat_c, mat_d))
# Read the number of total RNAs in each region if the bam file is given
if settings.total_RNA:
# prepare a dictionary of features
feat_dict = defaultdict(lambda: defaultdict(set))
for region1 in region_ints_as1:
for i in range(settings.seglen):
feat_dict[region1[2]+region1[1]][region1[0]+i].add(region1)
for region2 in region_ints_as2:
for i in range(settings.seglen):
feat_dict[region2[2]+region2[1]][region2[0]+i].add(region2)
feat_list = {}
for chrom, data in feat_dict.items():
maxpos = max(data.keys())
list_of_sets = []
for k in range(maxpos+1):
list_of_sets.append(list(data[k]))
feat_list[chrom] = list_of_sets
totRNA_counts = defaultdict(int)
for bamfile in settings.total_RNA.split(','):
totcounts = RILseq.count_features(
feat_list, pysam.Samfile(bamfile), 5,
rev=settings.total_reverse)
for k, v in totcounts.items():
totRNA_counts[k] += v
# Collect all the ratios between IP and total then choose the 90%
# percentile to avoid liers
max_IP_div_total_as1 = []
max_IP_div_total_as2 = []
for reg, counts in totRNA_counts.items():
if counts > settings.min_total_counts:
counts = float(counts+1)
# div_prod = max(region_ints_as1[reg]/counts, region_ints_as2[reg]/counts)
max_IP_div_total_as1.append(region_ints_as1[reg]/counts)
max_IP_div_total_as2.append(region_ints_as2[reg]/counts)
# mm1_sorted = sorted(max_IP_div_total_as1)
mm_sorted = sorted(max_IP_div_total_as2+max_IP_div_total_as1)
max_IP_div_total = mm_sorted[
int(len(mm_sorted)*settings.norm_percentile)]
sys.stderr.write("%f\n"%(max_IP_div_total))
else:
totRNA_counts = defaultdict(int)
max_IP_div_total = 0
# Read the additional data to decorate the results with
RILseq.report_interactions(
region_interactions, sys.stdout, interacting_regions, settings.seglen,
settings.ec_dir, settings.EC_chrlist, settings.refseq_dir,
settings.targets_file, settings.rep_table, settings.single_counts,
settings.shuffles, settings.RNAup_cmd, settings.servers,
settings.length, settings.est_utr_lens, settings.pad_seqs,
totRNA_counts, max_IP_div_total)
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
# Read the read names and positions
read_5ps = {}
read_3ps = {}
read_genes = {}
genome = {}
gsize = {}
for sr in SeqIO.parse(settings.genome, 'fasta'):
genome[sr.id] = sr.seq
gsize[sr.id] = len(sr.seq) # genome size dictionary - {chr:size}
if len(settings.BC_chrlist) >= 2:
chr_dict = dict(zip(
settings.BC_chrlist.split(',')[0::2],
settings.BC_chrlist.split(',')[1::2]))
# create dictionary of {'COLI-K12' : 'chr'}
else:
chr_dict = {}
if settings.summary: # only reads from the significant interactions in -s param, also can enter a specific gene.
sig_reads = RILseq.read_significant_reads(
settings.summary, chr_dict, gname=settings.gene_name)
for line in csv.reader(open(settings.list_reads), delimiter='\t'):
# skip single
if len(line) > 7 and line[7]=="single":
continue
if settings.summary:
if (int(line[4])-1, line[5], line[3]) not in\
sig_reads[(int(line[1])-1, line[2], line[0])]:
# skip if (coord_2, strand_2, chrom_2) not in the (coord_2, strand_2, chrom_2) of the significant reads.
continue
read_5ps[line[6]] = [int(line[1])-1, line[2], line[0]] # {read_id : [coord_1(0-based), strand_1, chrom_1]}
read_3ps[line[6]] = [int(line[4])-1, line[5], line[3]] # {read_id : [coord_2(0-based), strand_2, chrom_2]}
# read_genes[line[6]] = [line[0], line[1]]
# Read the bam files and return the long sequences
r1_seqs = {}
r2_seqs = {}
for bamfile in list(RILseq.flat_list(settings.bamfiles)): # flat multiple lists into one list.
r1s, r2s = get_reads_seqs(
pysam.AlignmentFile(bamfile, 'rb'), read_5ps.keys(), rev=settings.reverse)
r1_seqs.update(r1s)
r2_seqs.update(r2s)
# For each read find the overlap, if exists and find the fusion point
outer = csv.writer(sys.stdout, delimiter='\t')
print ('track name="%s" description="%s" visibility=4 itemRgb="On" useScore=0'%(
settings.track_name, settings.track_desc))
# Because I'm lazy, the code is written so r1 is the 3' end of the fragment
for rname in set(r2_seqs.keys()):
if rname in r1_seqs:
r2seq = r2_seqs[rname]
r1seq = r1_seqs[rname]
else: # single-end
r2seq = r2_seqs[rname]
r1seq = ''
# r2seq, r1seq are the sequences from the bam files for paired end
s1, overlap, s2 = find_overlap(r2seq, r1seq)
# print here
print(s1+overlap+s2)
print(read_5ps[rname][0])
print(read_3ps[rname][0])
side_5p_len = extend_alignment(
s1+overlap+s2, read_5ps[rname][0], 0, False, read_5ps[rname][1],
genome[read_5ps[rname][2]])
side_3p_len = extend_alignment(
s1+overlap+s2, 0, read_3ps[rname][0], True, read_3ps[rname][1],
genome[read_3ps[rname][2]])
# Write each of the sides to the output file
score=0
if settings.rand_score:
score=random.randint(0, 1000)
if read_5ps[rname][1] == '+':
gfrom = max(0, read_5ps[rname][0])
gto = min(gsize[read_5ps[rname][2]], read_5ps[rname][0]+side_5p_len)
outer.writerow([
read_5ps[rname][2], gfrom, gto, "%s_5p"%rname, score, '+',
gfrom, gto, settings.pos_first])
elif read_5ps[rname][1] == '-':
gfrom = max(0, read_5ps[rname][0]-side_5p_len+1)
gto = min(gsize[read_5ps[rname][2]], read_5ps[rname][0]+1)
outer.writerow([
read_5ps[rname][2], gfrom, gto, "%s_5p"%rname, score, '-',
gfrom, gto, settings.rev_first])
if read_3ps[rname][1] == '+':
gfrom = max(0, read_3ps[rname][0]-side_3p_len+1)
gto = min(gsize[read_3ps[rname][2]], read_3ps[rname][0]+1)
outer.writerow([
read_3ps[rname][2], gfrom, gto,"%s_3p"%rname, score, '+',
gfrom, gto, settings.pos_second])
elif read_3ps[rname][1] == '-':
gfrom = max(0, read_3ps[rname][0])
gto = min(gsize[read_3ps[rname][2]], read_3ps[rname][0]+side_3p_len)
outer.writerow([
read_3ps[rname][2], gfrom, gto, "%s_3p"%rname, score, '-',
gfrom, gto, settings.rev_second])
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
if len(settings.EC_chrlist) >= 2:
chr_dict = dict(zip(
settings.EC_chrlist.split(',')[0::2],
settings.EC_chrlist.split(',')[1::2]))
else:
chr_dict = {}
region_interactions, _, _, _=\
RILseq.read_reads_table(open(settings.reads_in), settings.region)
both_strs = defaultdict(lambda: defaultdict(int))
if settings.summary:
sig_reads = RILseq.read_significant_reads(
settings.summary, chr_dict)
if settings.known:
known_reads = defaultdict(list)
ptt_c, ptt_str = get_coords("%s.ptt.gz"%settings.refseq_dir)
rnt_c, rnt_str = get_coords("%s.rnt.gz"%settings.refseq_dir)
for line in open(settings.known):
spl = line.strip().split()
try:
scoor = rnt_c[spl[0]]
rcoor = ptt_c[spl[1]]
except KeyError:
pass
else:
for i in range(scoor[0], scoor[1]):
for j in range(rcoor[0], rcoor[1]):
known_reads[i].append(j)
known_reads[j].append(i)
for reg1 in region_interactions:
if reg1[2] != settings.chrn:
continue
for reg2 in region_interactions[reg1]:
if reg2[2] != settings.chrn:
continue
if settings.summary:
nsigs = 0
for r1, r2, in region_interactions[reg1][reg2]:
nsigs += int((r2, reg2[1], reg2[2]) in \
sig_reads[(r1, reg1[1], reg1[2])])
else:
nsigs = len(region_interactions[reg1][reg2])
both_strs[reg1[0]][reg2[0]] += nsigs
if settings.sRNAs:
from RILseq.ecocyc_parser import read_genes_data
uid_pos, uid_names, uid_tudata, sRNAs_list, other_RNAs_list, rRNAs = \
read_genes_data(settings.ec_dir)
sposs = set()
for g in sRNAs_list:
for i in range(uid_pos[g][1], uid_pos[g][2]):
sposs.add(i)
for r1 in both_strs:
for r2 in both_strs[r1]:
if both_strs[r1][r2] > settings.min_interactions:
color = 'thickness=%dp'%max(
int(log(both_strs[r1][r2])/log(10)),1)
if settings.sRNAs:
rset = set([i for i in range(r1, r1+settings.region)])
rset |= set([i for i in range(r2, r2+settings.region)])
if rset & sposs:
color = 'color=orange'
if settings.known:
for k in set(range(r1, r1+settings.region)) & set(known_reads.keys()):
if set(range(r2, r2+settings.region)) & set(known_reads[k]):
color = 'color=red'
sys.stdout.write('%s %d %d %s %d %d %s\n'%(
settings.print_chr, r1+1, r1+settings.region,
settings.print_chr, r2+1, r2+settings.region,
color))
return 0 # success
def main(argv=None):
settings = process_command_line(argv)
# Read the transcripts if given
if settings.transcripts:
trans_dict = RILseq.read_transcripts(settings.transcripts)
else:
trans_dict = None
# Get the ends of the reads from the bam files
# sys.stderr.write('%s\n'%str(settings.bamfiles))
if settings.all_reads:
try:
outall = open(settings.all_reads, 'w')
except IOError:
outall = None
elif settings.add_all_reads:
outall = sys.stdout
else:
outall = None
for bf in RILseq.flat_list(settings.bamfiles):
bfin = pysam.Samfile(bf)
outhead = bf.rsplit('.', 1)[0]
libname = outhead.rsplit('/', 1)[-1]
fsq1name = "%s/%s_ends_1.fastq" % (settings.dirout, libname)
fsq2name = "%s/%s_ends_2.fastq" % (settings.dirout, libname)
if settings.skip_mapping:
fsq1 = open(os.devnull, 'w')
fsq2 = fsq1
else:
fsq1 = open(fsq1name, 'w')
fsq2 = open(fsq2name, 'w')
single_mapped = RILseq.get_unmapped_reads(
bfin,
fsq1,
fsq2,
settings.length,
settings.maxG,
rev=settings.reverse_complement,
all_reads=True,
dust_thr=settings.dust_thr)
reads_in = []
# Map the fastq files to the genome
for fqname in (fsq1name, fsq2name):
bamheadname = fqname.rsplit('.', 1)[0].rsplit('/', 1)[-1]
if settings.skip_mapping:
bamname = "%s/%s.bam" % (settings.dirout, bamheadname)
else:
bamname = RILseq.run_bwa(settings.bwa_exec,
fqname,
None,
settings.dirout,
bamheadname,
settings.max_mismatches,
settings.genome_fasta,
settings.params_aln,
'',
settings.samse_params,
settings.samtools_cmd,
processors=settings.processors)
bamin = pysam.Samfile(bamname)
reads_in.append(
RILseq.read_bam_file(bamin, bamin.references,
settings.allowed_mismatches))
RILseq.write_reads_table(sys.stdout,
reads_in[0],
reads_in[1],
bfin.references,
settings.distance,
not settings.keep_circular,
trans_dict,
write_single=outall,
single_mapped=single_mapped,
max_NM=settings.allowed_mismatches)
return 0 # success