def main():
parser = argparse.ArgumentParser(description="Filter a genepred by transcript length")
parser.add_argument('input',help="Input '-' for STDOUT")
parser.add_argument('--min_length',type=int,help="Minimum transcript length")
parser.add_argument('--max_length',type=int,help="Maximum transcript length")
parser.add_argument('--names',help="filter on a name list")
parser.add_argument('--gene_names',help="filter on a gene name list")
parser.add_argument('-v','--invert',action='store_true',help='Invert search result')
args = parser.parse_args()
name_list = set()
gene_name_list = set()
if args.names:
with open(args.names) as inf:
for line in inf:
f = line.rstrip().split("\t")
name_list.add(f[0])
if args.gene_names:
with open(args.gene_names) as inf:
for line in inf:
f = line.rstrip().split("\t")
gene_name_list.add(f[0])
inf = sys.stdin
if args.input != '-':
inf = open(args.input)
for line in inf:
if re.match('^#',line): continue
is_good = True
g = GPD(line.rstrip())
tot = g.length()
if args.min_length:
if tot < args.min_length:
is_good = False
if args.max_length:
if tot > args.max_length:
is_good = False
if args.names:
if g.value('name') not in name_list:
is_good = False
if args.gene_names:
if g.value('gene_name') not in args.gene_name_list:
is_good = False
# If we are still here we can print
if not args.invert:
if is_good: print line.rstrip()
else:
if not is_good: print line.rstrip()
def main():
parser = argparse.ArgumentParser(description="For every genepred entry report its alignability",formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('input',help="Genepred can be gzipped or - for STDIN")
parser.add_argument('-r','--reference',required=True,help="Reference fasta")
parser.add_argument('-k','--fragment_size',default=100,type=int,help="Fragment size to try to align")
parser.add_argument('-x','--hisat_index',required=True,help="HISAT index base name")
parser.add_argument('--threads',type=int,default=cpu_count(),help="number of threads")
parser.add_argument('--type',choices=['mean','median'],default='mean',help="How to bring together overlapping reads")
parser.add_argument('--perbase',action='store_true')
parser.add_argument('--output','-o',help="output file or leave unset for STDOUT")
args = parser.parse_args()
if args.input=='-': args.input=sys.stdin
elif re.search('\.gz$',args.input):
args.input = gzip.open(args.input)
else: args.input = open(args.input)
udir = os.path.dirname(os.path.realpath(__file__))
cmd2 = udir+'/genepred_counts_to_mappability.py -'
cmd2 += ' --threads '+str(args.threads)
cmd2 += ' -k '+str(args.fragment_size)
if args.perbase: cmd2 += ' --perbase'
if args.output: cmd2 += ' --output '+args.output
if args.type: cmd2 += ' --type '+args.type
p2 = Popen(cmd2.split(),stdin=PIPE)
ref = read_fasta_into_hash(args.reference)
cmd1 = 'hisat -x '+args.hisat_index+' -U - -f --reorder -p '+str(args.threads)
p1 = Popen(cmd1.split(),stdin=PIPE,stdout=p2.stdin,stderr=null)
#p1 = Popen(cmd1.split(),stdin=PIPE,stdout=p2.stdin)
line_number = 0
for line in args.input:
line_number +=1
gpd = GPD(line.rstrip())
#print gpd.entry['name']
#print gpd.length()
if gpd.length() < args.fragment_size: continue
seq = gpd.get_sequence(ref)
for i in range(0,len(seq)-args.fragment_size+1):
info = gpd.value('name')+"\t"+gpd.value('gene_name')+"\t"+str(line_number)+"\t"+str(len(seq))+"\t"+str(i)
einfo = encode_name(info)
p1.stdin.write('>'+einfo+"\n")
p1.stdin.write(seq[i:i+args.fragment_size]+"\n")
p1.communicate()
p2.communicate()
def main():
parser = argparse.ArgumentParser()
parser.add_argument('gpd_input')
parser.add_argument('bam_input')
parser.add_argument('--intergenic_buffer',default=10000,type=int)
parser.add_argument('--window_size',default=10000,type=int)
parser.add_argument('--bin_size',default=1000,type=int)
parser.add_argument('--use_off_regions',action='store_true',help="Use a region even if there is no reads mapped to it.")
parser.add_argument('--get_exons',action='store_true')
args = parser.parse_args()
chr_beds = {}
gene_beds = []
exon_beds = []
sys.stderr.write("Reading genepred file\n")
asum = 0
atot = 0
with open(args.gpd_input) as inf:
for line in inf:
g = GenePredEntry(line)
asum += g.length()
atot += 1
grng = g.get_bed()
grng.direction = None
if grng.chr not in chr_beds:
chr_beds[grng.chr] = grng.copy()
chr_beds[grng.chr] = chr_beds[grng.chr].merge(grng)
gene_beds.append(grng)
for i in range(0,g.get_exon_count()):
erng = Bed(g.value('chrom'),g.value('exonStarts')[i],g.value('exonEnds')[i])
exon_beds.append(erng)
avglen = float(asum)/float(atot)
sys.stderr.write("Sorting gene bed\n")
gene_beds = sort_ranges(gene_beds)
gene_beds = merge_ranges(gene_beds,already_sorted=True)
sys.stderr.write("Sorting chromosome beds\n")
chr_beds = sort_ranges([chr_beds[x] for x in chr_beds.keys()])
sys.stderr.write("Sorting exon beds\n")
exon_beds = sort_ranges(exon_beds)
sys.stderr.write("Get padded genes\n")
padded_gene_beds = pad_ranges(gene_beds,args.intergenic_buffer,chr_beds)
padded_gene_beds = merge_ranges(padded_gene_beds,already_sorted=True)
sys.stderr.write("Get intergenic regions\n")
intergenic_beds = subtract_ranges(chr_beds,padded_gene_beds,already_sorted=True)
intergenic_beds = merge_ranges(intergenic_beds,already_sorted=True)
intergenic_beds = window_break(intergenic_beds,args.window_size)
#for i in intergenic_beds: print i.get_range_string()
sys.stderr.write("Get merged exons\n")
exon_beds = merge_ranges(exon_beds)
sys.stderr.write("Get introns\n")
intron_beds = subtract_ranges(gene_beds,exon_beds,already_sorted=True)
intron_beds = merge_ranges(intron_beds,already_sorted=True)
intron_beds = window_break(intron_beds,args.window_size)
sys.stderr.write("Going through short reads\n")
cmd = "sam_to_bed_depth.py "+args.bam_input
p = Popen(cmd.split(),stdout=PIPE)
for x in intron_beds: x.set_payload([]) # payloads are read depths
for x in intergenic_beds: x.set_payload([]) # payloads are read depths
for x in exon_beds: x.set_payload([]) # payloads are read depths
introndepth = []
intergenicdepth = []
exondepth = []
pseudoreadcount = 0
if not args.get_exons: exon_beds = []
section_count = 0
while True:
section_count += 1
line = p.stdout.readline()
if not line: break
f = line.split("\t")
depth = int(f[3])
curr = Bed(f[0],int(f[1]),int(f[2]))
if section_count %100==0: sys.stderr.write(curr.get_range_string()+" \r")
pseudoreadcount += depth
if len(exon_beds) > 0:
while curr.cmp(exon_beds[0]) > 0 and len(exon_beds) > 0: # we've passed the region
v = exon_beds.pop(0)
if len(v.get_payload()) == 0 and not args.use_off_regions: continue
av = average(v)
exondepth.append(av)
#print str(av)+" exonic "+v.get_range_string()
c = curr.cmp(exon_beds[0])
if c == 0: # overlaps with intron
size = curr.overlap_size(exon_beds[0])
for i in range(0,size): exon_beds[0].get_payload().append(depth)
if len(intron_beds) > 0:
while curr.cmp(intron_beds[0]) > 0 and len(intron_beds) > 0: # we've passed the region
v = intron_beds.pop(0)
if len(v.get_payload()) == 0 and not args.use_off_regions: continue
av = average(v)
introndepth.append(av)
#print str(av)+" intronic "+v.get_range_string()
c = curr.cmp(intron_beds[0])
if c == 0: # overlaps with intron
size = curr.overlap_size(intron_beds[0])
for i in range(0,size): intron_beds[0].get_payload().append(depth)
if len(intergenic_beds) > 0:
while curr.cmp(intergenic_beds[0]) > 0 and len(intergenic_beds) > 0: # we've passed the region
v = intergenic_beds.pop(0)
if len(v.get_payload()) == 0 and not args.use_off_regions: continue
av = average(v)
intergenicdepth.append(av)
display(curr,introndepth,intergenicdepth,pseudoreadcount,avglen)
#print str(av)+" intergenic "+v.get_range_string()
c = curr.cmp(intergenic_beds[0])
if c == 0: # overlaps with intron
size = curr.overlap_size(intergenic_beds[0])
for i in range(0,size): intergenic_beds[0].get_payload().append(depth)
#if c > 0: # we passed the intron
# v = intergenic_beds.pop(0)
# av = average(v)
# intergenicdepth.append(av)
# print str(av)+" intergenic "+v.get_range_string()
if args.use_off_regions:
for x in exon_beds: introndepth.append(average(x.get_payload()))
for x in intron_beds: introndepth.append(average(x.get_payload()))
for x in intergenic_beds: intergenicdepth.append(average(x.get_payload()))
p.communicate()
def main():
parser = argparse.ArgumentParser()
parser.add_argument('gpd_input')
parser.add_argument('bam_input')
parser.add_argument('--intergenic_buffer', default=10000, type=int)
parser.add_argument('--window_size', default=10000, type=int)
parser.add_argument('--bin_size', default=1000, type=int)
parser.add_argument(
'--use_off_regions',
action='store_true',
help="Use a region even if there is no reads mapped to it.")
parser.add_argument('--get_exons', action='store_true')
args = parser.parse_args()
chr_beds = {}
gene_beds = []
exon_beds = []
sys.stderr.write("Reading genepred file\n")
asum = 0
atot = 0
with open(args.gpd_input) as inf:
for line in inf:
g = GenePredEntry(line)
asum += g.length()
atot += 1
grng = g.get_bed()
grng.direction = None
if grng.chr not in chr_beds:
chr_beds[grng.chr] = grng.copy()
chr_beds[grng.chr] = chr_beds[grng.chr].merge(grng)
gene_beds.append(grng)
for i in range(0, g.get_exon_count()):
erng = Bed(g.value('chrom'),
g.value('exonStarts')[i],
g.value('exonEnds')[i])
exon_beds.append(erng)
avglen = float(asum) / float(atot)
sys.stderr.write("Sorting gene bed\n")
gene_beds = sort_ranges(gene_beds)
gene_beds = merge_ranges(gene_beds, already_sorted=True)
sys.stderr.write("Sorting chromosome beds\n")
chr_beds = sort_ranges([chr_beds[x] for x in chr_beds.keys()])
sys.stderr.write("Sorting exon beds\n")
exon_beds = sort_ranges(exon_beds)
sys.stderr.write("Get padded genes\n")
padded_gene_beds = pad_ranges(gene_beds, args.intergenic_buffer, chr_beds)
padded_gene_beds = merge_ranges(padded_gene_beds, already_sorted=True)
sys.stderr.write("Get intergenic regions\n")
intergenic_beds = subtract_ranges(chr_beds,
padded_gene_beds,
already_sorted=True)
intergenic_beds = merge_ranges(intergenic_beds, already_sorted=True)
intergenic_beds = window_break(intergenic_beds, args.window_size)
#for i in intergenic_beds: print i.get_range_string()
sys.stderr.write("Get merged exons\n")
exon_beds = merge_ranges(exon_beds)
sys.stderr.write("Get introns\n")
intron_beds = subtract_ranges(gene_beds, exon_beds, already_sorted=True)
intron_beds = merge_ranges(intron_beds, already_sorted=True)
intron_beds = window_break(intron_beds, args.window_size)
sys.stderr.write("Going through short reads\n")
cmd = "sam_to_bed_depth.py " + args.bam_input
p = Popen(cmd.split(), stdout=PIPE)
for x in intron_beds:
x.set_payload([]) # payloads are read depths
for x in intergenic_beds:
x.set_payload([]) # payloads are read depths
for x in exon_beds:
x.set_payload([]) # payloads are read depths
introndepth = []
intergenicdepth = []
exondepth = []
pseudoreadcount = 0
if not args.get_exons: exon_beds = []
section_count = 0
while True:
section_count += 1
line = p.stdout.readline()
if not line: break
f = line.split("\t")
depth = int(f[3])
curr = Bed(f[0], int(f[1]), int(f[2]))
if section_count % 100 == 0:
sys.stderr.write(curr.get_range_string() + " \r")
pseudoreadcount += depth
if len(exon_beds) > 0:
while curr.cmp(exon_beds[0]) > 0 and len(
exon_beds) > 0: # we've passed the region
v = exon_beds.pop(0)
if len(v.get_payload()) == 0 and not args.use_off_regions:
continue
av = average(v)
exondepth.append(av)
#print str(av)+" exonic "+v.get_range_string()
c = curr.cmp(exon_beds[0])
if c == 0: # overlaps with intron
size = curr.overlap_size(exon_beds[0])
for i in range(0, size):
exon_beds[0].get_payload().append(depth)
if len(intron_beds) > 0:
while curr.cmp(intron_beds[0]) > 0 and len(
intron_beds) > 0: # we've passed the region
v = intron_beds.pop(0)
if len(v.get_payload()) == 0 and not args.use_off_regions:
continue
av = average(v)
introndepth.append(av)
#print str(av)+" intronic "+v.get_range_string()
c = curr.cmp(intron_beds[0])
if c == 0: # overlaps with intron
size = curr.overlap_size(intron_beds[0])
for i in range(0, size):
intron_beds[0].get_payload().append(depth)
if len(intergenic_beds) > 0:
while curr.cmp(intergenic_beds[0]) > 0 and len(
intergenic_beds) > 0: # we've passed the region
v = intergenic_beds.pop(0)
if len(v.get_payload()) == 0 and not args.use_off_regions:
continue
av = average(v)
intergenicdepth.append(av)
display(curr, introndepth, intergenicdepth, pseudoreadcount,
avglen)
#print str(av)+" intergenic "+v.get_range_string()
c = curr.cmp(intergenic_beds[0])
if c == 0: # overlaps with intron
size = curr.overlap_size(intergenic_beds[0])
for i in range(0, size):
intergenic_beds[0].get_payload().append(depth)
#if c > 0: # we passed the intron
# v = intergenic_beds.pop(0)
# av = average(v)
# intergenicdepth.append(av)
# print str(av)+" intergenic "+v.get_range_string()
if args.use_off_regions:
for x in exon_beds:
introndepth.append(average(x.get_payload()))
for x in intron_beds:
introndepth.append(average(x.get_payload()))
for x in intergenic_beds:
intergenicdepth.append(average(x.get_payload()))
p.communicate()
def main():
parser = argparse.ArgumentParser(
description="For every genepred entry report its alignability",
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('input', help="Genepred can be gzipped or - for STDIN")
parser.add_argument('-r',
'--reference',
required=True,
help="Reference fasta")
parser.add_argument('-k',
'--fragment_size',
default=100,
type=int,
help="Fragment size to try to align")
parser.add_argument('-x',
'--hisat_index',
required=True,
help="HISAT index base name")
parser.add_argument('--threads',
type=int,
default=cpu_count(),
help="number of threads")
parser.add_argument('--type',
choices=['mean', 'median'],
default='mean',
help="How to bring together overlapping reads")
parser.add_argument('--perbase', action='store_true')
parser.add_argument('--output',
'-o',
help="output file or leave unset for STDOUT")
args = parser.parse_args()
if args.input == '-': args.input = sys.stdin
elif re.search('\.gz$', args.input):
args.input = gzip.open(args.input)
else:
args.input = open(args.input)
udir = os.path.dirname(os.path.realpath(__file__))
cmd2 = udir + '/genepred_counts_to_mappability.py -'
cmd2 += ' --threads ' + str(args.threads)
cmd2 += ' -k ' + str(args.fragment_size)
if args.perbase: cmd2 += ' --perbase'
if args.output: cmd2 += ' --output ' + args.output
if args.type: cmd2 += ' --type ' + args.type
p2 = Popen(cmd2.split(), stdin=PIPE)
ref = read_fasta_into_hash(args.reference)
cmd1 = 'hisat -x ' + args.hisat_index + ' -U - -f --reorder -p ' + str(
args.threads)
p1 = Popen(cmd1.split(), stdin=PIPE, stdout=p2.stdin, stderr=null)
#p1 = Popen(cmd1.split(),stdin=PIPE,stdout=p2.stdin)
line_number = 0
for line in args.input:
line_number += 1
gpd = GPD(line.rstrip())
#print gpd.entry['name']
#print gpd.length()
if gpd.length() < args.fragment_size: continue
seq = gpd.get_sequence(ref)
for i in range(0, len(seq) - args.fragment_size + 1):
info = gpd.value('name') + "\t" + gpd.value(
'gene_name') + "\t" + str(line_number) + "\t" + str(
len(seq)) + "\t" + str(i)
einfo = encode_name(info)
p1.stdin.write('>' + einfo + "\n")
p1.stdin.write(seq[i:i + args.fragment_size] + "\n")
p1.communicate()
p2.communicate()
