def _bigwig_extractor(datafile, intervals, **kwargs):
width = intervals[0].stop - intervals[0].start
data = np.zeros((len(intervals), 1, 1, width))
wWigIO.open(datafile)
for index, interval in enumerate(intervals):
wWigIO.getData(datafile, interval.chrom, interval.start, interval.stop,
data[index, 0, 0, :])
wWigIO.close(datafile)
return data
def extract_bigwig_to_npy(bigwig, output_dir, dtype=np.float32):
wWigIO.open(bigwig)
chrom_sizes = wWigIO.getChromSize(bigwig)
file_shapes = {}
for chrom, size in zip(*chrom_sizes):
data = np.empty(size)
wWigIO.getData(bigwig, chrom, 0, size, data)
np.save('{}.npy'.format(os.path.join(output_dir, chrom)),
data.astype(dtype))
file_shapes[chrom] = data.shape
wWigIO.close(bigwig)
with open(os.path.join(output_dir, 'metadata.json'), 'w') as fp:
json.dump(
{
'file_shapes': file_shapes,
'type': 'array',
'source': bigwig
}, fp)
def getChromSizesWigIO(self):
wWigIO.open(self.path)
out = wWigIO.getChromSize(self.path)
out = dict(zip(out[0], out[1]))
wWigIO.close(self.path)
return out
# ------------------------------------
# Misc functions
# ------------------------------------
# ------------------------------------
# Classes
# ------------------------------------
# ------------------------------------
# Main
# ------------------------------------
if __name__=="__main__":
# get information from bigwig file
wWigIO.open('test.bw')
chroms = wWigIO.getChromSize('test.bw')
wigs = wWigIO.getIntervals('test.bw', 'chr1', 10, 200)
wWigIO.close('test.bw')
print wigs
# bigwig -> wig
wWigIO.bigWigToWig('test.bw','test.wig')
# write the chrom sizes into test.sizes
with open('test.sizes','w') as fh:
for chrom in chroms:
print >>fh, chrom+"\t"+str(chroms[chrom])
# wig -> bigwig
wWigIO.wigToBigWig('test.wig','test.sizes','test2.bw')
def main(row_ME, phylop_primates): wWigIO.open(phylop_vertebrates) wWigIO.open(phylop_primates)
def main(sim_fastq, U2_GTAG_5_file, U2_GTAG_3_file, phylop_vertebrates,
phylop_primates, exon_scores):
MEs = set([])
wWigIO.open(phylop_vertebrates)
wWigIO.open(phylop_primates)
U2_GTAG_5 = PWM_to_dict(U2_GTAG_5_file)
U2_GTAG_3 = PWM_to_dict(U2_GTAG_3_file)
U2_GTAG_5_max_score = 0
U2_GTAG_3_max_score = 0
for index in range(13):
U2_GTAG_5_max_score += max(U2_GTAG_5['A'][index],
U2_GTAG_5['C'][index],
U2_GTAG_5['T'][index],
U2_GTAG_5['G'][index])
for index in range(17):
U2_GTAG_3_max_score += max(U2_GTAG_3['A'][index],
U2_GTAG_3['C'][index],
U2_GTAG_3['T'][index],
U2_GTAG_3['G'][index])
TOTAL_U2_max_score = U2_GTAG_5_max_score + U2_GTAG_3_max_score
gencode_U2_scores = []
gencode_mean_conservation_vertebrates = []
gencode_mean_conservation_primates = []
for row in csv.reader(open(exon_scores), delimiter=' '):
chr, estart, eend, strand, U2_score, mean_conservation_vertebrates, mean_conservation_primates = row
gencode_U2_scores.append(float(U2_score))
gencode_mean_conservation_vertebrates.append(
float(mean_conservation_vertebrates))
gencode_mean_conservation_primates.append(
float(mean_conservation_primates))
for row in csv.reader(open(sim_fastq), delimiter='\t'):
if row[0][0] == "@":
SJ, ME_seq, estart, eend, total_coverage, n = row[0].split("_")
len_ME = len(ME_seq)
SJ = SJ[1:]
SJ_chr, SJ_istart, SJ_iend = re.findall(r"[\w']+", SJ)
SJ_len = int(SJ_iend) - int(SJ_istart)
Kmer = SJ_len - (len_ME + 1)
P_ME = 1 - (1 - (float(1) / float(4**len_ME + 4)))**Kmer
strand = "+"
if "-" in SJ:
strand = "-"
estart = int(estart)
eend = int(eend)
MEs.add((SJ_chr, strand, estart, eend, P_ME))
for m in MEs:
chr, strand, estart, eend, P_ME = m
estart, eend = sorted([estart, eend])
E5 = str(Genome[chr][estart - 14:estart + 3]).upper()
E3 = str(Genome[chr][eend - 3:eend + 10]).upper()
if strand == "-":
E5 = str(Genome[chr][eend - 3:eend +
14].reverse_complement()).upper()
E3 = str(Genome[chr][estart - 10:estart +
3].reverse_complement()).upper()
U2_score = 0
i = 0
for N in E5:
if N != "N":
U2_score += U2_GTAG_3[N][i]
i += 1
i = 0
for N in E3:
if N != "N":
U2_score += U2_GTAG_5[N][i]
i += 1
U2_score = percent(U2_score, TOTAL_U2_max_score)
conservation_vertebrates = wWigIO.getIntervals(phylop_vertebrates, chr,
estart - 2, eend + 2)
conservation_primates = wWigIO.getIntervals(phylop_primates, chr,
estart - 2, eend + 2)
mean_conservation_vertebrates = 0
mean_conservation_primates = 0
for i in conservation_vertebrates:
mean_conservation_vertebrates += i[2]
try:
mean_conservation_vertebrates = mean_conservation_vertebrates / len(
conservation_vertebrates)
except ZeroDivisionError:
pass
for i in conservation_primates:
mean_conservation_primates += i[2]
try:
mean_conservation_primates = mean_conservation_primates / len(
conservation_primates)
except ZeroDivisionError:
pass
ME_percentil_U2_score = stats.percentileofscore(
gencode_U2_scores, U2_score)
ME_percentil_mean_conservation_vertebrates = stats.percentileofscore(
gencode_mean_conservation_vertebrates, mean_conservation_primates)
ME_percentil_mean_conservation_primates = stats.percentileofscore(
gencode_mean_conservation_primates, mean_conservation_vertebrates)
overall_score = P_ME * (1 - ME_percentil_U2_score / 100) * (
1 - ME_percentil_mean_conservation_vertebrates / 100)
if ME_percentil_mean_conservation_primates > ME_percentil_mean_conservation_vertebrates:
overall_score = P_ME * (1 - ME_percentil_U2_score / 100) * (
1 - ME_percentil_mean_conservation_primates / 100)
#print chr, estart, eend, strand, U2_score, mean_conservation_vertebrates, mean_conservation_primates
print chr, estart, eend, strand, U2_score, ME_percentil_U2_score, mean_conservation_vertebrates, ME_percentil_mean_conservation_vertebrates, mean_conservation_primates, ME_percentil_mean_conservation_primates, P_ME, overall_score
def main(gencode_bed, U2_GTAG_5_file, U2_GTAG_3_file, phylop_vertebrates,
phylop_primates):
wWigIO.open(phylop_vertebrates)
wWigIO.open(phylop_primates)
U2_GTAG_5 = PWM_to_dict(U2_GTAG_5_file)
U2_GTAG_3 = PWM_to_dict(U2_GTAG_3_file)
U2_GTAG_5_max_score = 0
U2_GTAG_3_max_score = 0
for index in range(13):
U2_GTAG_5_max_score += max(U2_GTAG_5['A'][index],
U2_GTAG_5['C'][index],
U2_GTAG_5['T'][index],
U2_GTAG_5['G'][index])
for index in range(17):
U2_GTAG_3_max_score += max(U2_GTAG_3['A'][index],
U2_GTAG_3['C'][index],
U2_GTAG_3['T'][index],
U2_GTAG_3['G'][index])
TOTAL_U2_max_score = U2_GTAG_5_max_score + U2_GTAG_3_max_score
exons = set([])
for row in csv.reader(open(gencode_bed), delimiter='\t'):
csv.field_size_limit(1000000000)
qstarts = map(int, row[11].strip(",").split(","))
blocksizes = map(int, row[10].strip(",").split(","))
start = int(row[1])
strand = row[5]
bn = int(row[9])
chr = row[0]
for q1, b in zip(qstarts[1:-1], blocksizes[1:-1]):
estart = start + q1
eend = start + q1 + b
E5 = str(Genome[chr][estart - 14:estart + 3]).upper()
E3 = str(Genome[chr][eend - 3:eend + 10]).upper()
if strand == "-":
E5 = str(Genome[chr][eend - 3:eend +
14].reverse_complement()).upper()
E3 = str(Genome[chr][estart - 10:estart +
3].reverse_complement()).upper()
U2_score = 0
i = 0
for N in E5:
U2_score += U2_GTAG_3[N][i]
i += 1
i = 0
for N in E3:
U2_score += U2_GTAG_5[N][i]
i += 1
U2_score = percent(U2_score, TOTAL_U2_max_score)
if E5[-5:-3] == "AG" and E3[3:5] == "GT":
exons.add((chr, estart, eend, strand, U2_score))
# if " ".join([chr, estart, eend]) == "chr17 26597935 26598725":
# print
for e in exons:
chr, estart, eend, strand, U2_score = e
conservation_vertebrates = wWigIO.getIntervals(phylop_vertebrates, chr,
estart - 2, eend + 2)
conservation_primates = wWigIO.getIntervals(phylop_primates, chr,
estart - 2, eend + 2)
mean_conservation_vertebrates = 0
mean_conservation_primates = 0
for i in conservation_vertebrates:
mean_conservation_vertebrates += i[2]
try:
mean_conservation_vertebrates = mean_conservation_vertebrates / len(
conservation_vertebrates)
except ZeroDivisionError:
pass
for i in conservation_primates:
mean_conservation_primates += i[2]
try:
mean_conservation_primates = mean_conservation_primates / len(
conservation_primates)
except ZeroDivisionError:
pass
print chr, estart, eend, strand, U2_score, mean_conservation_vertebrates, mean_conservation_primates
def __init__(self, fname):
''' Open BigWig file. '''
self.fname = fname
wWigIO.open(self.fname)
import os, sys
import numpy as np
import scipy.stats
import tabix
import wWigIO
import bw_bin
def show_help():
print >> sys.stderr, "\n\tpython ", sys.argv[
0], " /datd/huboqiang/test_NOM/02.SingleC/mESC_gF28_1/singleC/chr10.ACG.TCG.bed.gz COUNT_U COUNT_M 15"
in_bw = "/datd/huboqiang/test_NOM/mESC_nuc.sort.bw"
wWigIO.open(in_bw)
class NDR(object):
def __init__(self,
chrom,
count_u,
count_m,
tb_file,
cutoff=1e-5,
depth=3,
bin_len=40,
step_len=20,
dist_len=140):
self.chrom = chrom
self.bin_len = bin_len
def __init__(self,fname):
''' Open BigWig file. '''
self.fname=fname
wWigIO.open(self.fname)
# ------------------------------------
# Misc functions
# ------------------------------------
# ------------------------------------
# Classes
# ------------------------------------
# ------------------------------------
# Main
# ------------------------------------
if __name__ == "__main__":
# get information from bigwig file
wWigIO.open('test.bw')
chroms = wWigIO.getChromSize('test.bw')
wigs = wWigIO.getIntervals('test.bw', 'chr1', 10, 200)
wWigIO.close('test.bw')
print wigs
# bigwig -> wig
wWigIO.bigWigToWig('test.bw', 'test.wig')
# write the chrom sizes into test.sizes
with open('test.sizes', 'w') as fh:
for chrom in chroms:
print >> fh, chrom + "\t" + str(chroms[chrom])
# wig -> bigwig
wWigIO.wigToBigWig('test.wig', 'test.sizes', 'test2.bw')
def main(sim_fastq, U2_GTAG_5_file, U2_GTAG_3_file, phylop_vertebrates,
phylop_primates):
MEs = set([])
wWigIO.open(phylop_vertebrates)
wWigIO.open(phylop_primates)
U2_GTAG_5 = PWM_to_dict(U2_GTAG_5_file)
U2_GTAG_3 = PWM_to_dict(U2_GTAG_3_file)
U2_GTAG_5_max_score = 0
U2_GTAG_3_max_score = 0
for index in range(13):
U2_GTAG_5_max_score += max(U2_GTAG_5['A'][index],
U2_GTAG_5['C'][index],
U2_GTAG_5['T'][index],
U2_GTAG_5['G'][index])
for index in range(17):
U2_GTAG_3_max_score += max(U2_GTAG_3['A'][index],
U2_GTAG_3['C'][index],
U2_GTAG_3['T'][index],
U2_GTAG_3['G'][index])
TOTAL_U2_max_score = U2_GTAG_5_max_score + U2_GTAG_3_max_score
for row in csv.reader(open(sim_fastq), delimiter='\t'):
chr, estart, eend, exon, exon_len, strand = row
estart = int(estart)
eend = int(eend)
# if row[0][0]=="@":
# SJ, ME_seq, estart, eend, total_coverage, n = row[0].split("_")
# len_ME = len(ME_seq)
# SJ = SJ[1:]
# SJ_chr, SJ_istart, SJ_iend = re.findall(r"[\w']+", SJ)
# SJ_len = int(SJ_iend) - int(SJ_istart)
# Kmer = SJ_len - (len_ME+1)
# P_ME = 1 - ( 1 - (float(1)/float(4**len_ME+4)))**Kmer
# strand = "+"
# if "-" in SJ:
# strand = "-"
# estart = int(estart)
# eend = int(eend)
# MEs.add((SJ_chr, strand, estart, eend, P_ME))
# for m in MEs:
# chr, strand, estart, eend, P_ME = m
# estart, eend = sorted([estart, eend])
E5 = str(Genome[chr][estart - 14:estart + 3]).upper()
E3 = str(Genome[chr][eend - 3:eend + 10]).upper()
if strand == "-":
E5 = str(Genome[chr][eend - 3:eend +
14].reverse_complement()).upper()
E3 = str(Genome[chr][estart - 10:estart +
3].reverse_complement()).upper()
E5 = E5[:-5] + "AG" + E5[-3:]
E3 = E3[:3] + "GT" + E3[5:]
U2_score = 0
ME5_U2_score = 0
ME3_U2_score = 0
i = 0
for N in E5:
if N != "N":
U2_score += U2_GTAG_3[N][i]
ME5_U2_score += U2_GTAG_3[N][i]
i += 1
i = 0
for N in E3:
if N != "N":
U2_score += U2_GTAG_5[N][i]
ME3_U2_score += U2_GTAG_5[N][i]
i += 1
ME3_U2_score = percent(ME3_U2_score, U2_GTAG_5_max_score)
ME5_U2_score = percent(ME5_U2_score, U2_GTAG_3_max_score)
U2_score = percent(U2_score, TOTAL_U2_max_score)
conservation_vertebrates = wWigIO.getIntervals(phylop_vertebrates, chr,
estart - 2, eend + 2)
conservation_primates = wWigIO.getIntervals(phylop_primates, chr,
estart - 2, eend + 2)
mean_conservation_vertebrates = 0
mean_conservation_primates = 0
for i in conservation_vertebrates:
mean_conservation_vertebrates += i[2]
try:
mean_conservation_vertebrates = mean_conservation_vertebrates / len(
conservation_vertebrates)
except ZeroDivisionError:
pass
for i in conservation_primates:
mean_conservation_primates += i[2]
try:
mean_conservation_primates = mean_conservation_primates / len(
conservation_primates)
except ZeroDivisionError:
pass
#print chr, estart, eend, strand, U2_score, mean_conservation_vertebrates, mean_conservation_primates
print chr, estart, eend, strand, U2_score, ME5_U2_score, ME3_U2_score, mean_conservation_vertebrates, mean_conservation_primates
def main(row_ME, phylop_primates):
wWigIO.open(phylop_vertebrates)
wWigIO.open(phylop_primates)
def main(sim_fastq, U2_GTAG_5_file, U2_GTAG_3_file, phylop_vertebrates, phylop_primates):
MEs = set([])
wWigIO.open(phylop_vertebrates)
wWigIO.open(phylop_primates)
U2_GTAG_5 = PWM_to_dict(U2_GTAG_5_file)
U2_GTAG_3 = PWM_to_dict(U2_GTAG_3_file)
U2_GTAG_5_max_score = 0
U2_GTAG_3_max_score = 0
for index in range(13):
U2_GTAG_5_max_score += max(U2_GTAG_5['A'][index], U2_GTAG_5['C'][index], U2_GTAG_5['T'][index], U2_GTAG_5['G'][index])
for index in range(17):
U2_GTAG_3_max_score += max(U2_GTAG_3['A'][index], U2_GTAG_3['C'][index], U2_GTAG_3['T'][index], U2_GTAG_3['G'][index])
TOTAL_U2_max_score = U2_GTAG_5_max_score + U2_GTAG_3_max_score
for row in csv.reader(open(sim_fastq), delimiter = '\t'):
chr, estart, eend, exon, exon_len, strand = row
estart = int(estart)
eend = int(eend)
# if row[0][0]=="@":
# SJ, ME_seq, estart, eend, total_coverage, n = row[0].split("_")
# len_ME = len(ME_seq)
# SJ = SJ[1:]
# SJ_chr, SJ_istart, SJ_iend = re.findall(r"[\w']+", SJ)
# SJ_len = int(SJ_iend) - int(SJ_istart)
# Kmer = SJ_len - (len_ME+1)
# P_ME = 1 - ( 1 - (float(1)/float(4**len_ME+4)))**Kmer
# strand = "+"
# if "-" in SJ:
# strand = "-"
# estart = int(estart)
# eend = int(eend)
# MEs.add((SJ_chr, strand, estart, eend, P_ME))
# for m in MEs:
# chr, strand, estart, eend, P_ME = m
# estart, eend = sorted([estart, eend])
E5 = str(Genome[chr][estart-14:estart+3]).upper()
E3 = str(Genome[chr][eend-3:eend+10]).upper()
if strand == "-":
E5 = str(Genome[chr][eend-3:eend+14].reverse_complement()).upper()
E3 = str(Genome[chr][estart-10:estart+3].reverse_complement()).upper()
E5 = E5[:-5] + "AG" + E5[-3:]
E3 = E3[:3] + "GT" + E3[5:]
U2_score = 0
ME5_U2_score = 0
ME3_U2_score = 0
i = 0
for N in E5:
if N!="N":
U2_score += U2_GTAG_3[N][i]
ME5_U2_score += U2_GTAG_3[N][i]
i += 1
i = 0
for N in E3:
if N!="N":
U2_score += U2_GTAG_5[N][i]
ME3_U2_score += U2_GTAG_5[N][i]
i += 1
ME3_U2_score = percent(ME3_U2_score, U2_GTAG_5_max_score)
ME5_U2_score = percent(ME5_U2_score, U2_GTAG_3_max_score)
U2_score = percent(U2_score, TOTAL_U2_max_score)
conservation_vertebrates = wWigIO.getIntervals(phylop_vertebrates, chr, estart-2, eend+2)
conservation_primates = wWigIO.getIntervals(phylop_primates, chr, estart-2, eend+2)
mean_conservation_vertebrates = 0
mean_conservation_primates = 0
for i in conservation_vertebrates:
mean_conservation_vertebrates += i[2]
try:
mean_conservation_vertebrates = mean_conservation_vertebrates/len(conservation_vertebrates)
except ZeroDivisionError:
pass
for i in conservation_primates:
mean_conservation_primates += i[2]
try:
mean_conservation_primates = mean_conservation_primates/len(conservation_primates)
except ZeroDivisionError:
pass
#print chr, estart, eend, strand, U2_score, mean_conservation_vertebrates, mean_conservation_primates
print chr, estart, eend, strand, U2_score, ME5_U2_score, ME3_U2_score, mean_conservation_vertebrates, mean_conservation_primates
def main(gencode_bed, U2_GTAG_5_file, U2_GTAG_3_file, phylop_vertebrates, phylop_primates):
wWigIO.open(phylop_vertebrates)
wWigIO.open(phylop_primates)
U2_GTAG_5 = PWM_to_dict(U2_GTAG_5_file)
U2_GTAG_3 = PWM_to_dict(U2_GTAG_3_file)
U2_GTAG_5_max_score = 0
U2_GTAG_3_max_score = 0
for index in range(13):
U2_GTAG_5_max_score += max(U2_GTAG_5['A'][index], U2_GTAG_5['C'][index], U2_GTAG_5['T'][index], U2_GTAG_5['G'][index])
for index in range(17):
U2_GTAG_3_max_score += max(U2_GTAG_3['A'][index], U2_GTAG_3['C'][index], U2_GTAG_3['T'][index], U2_GTAG_3['G'][index])
TOTAL_U2_max_score = U2_GTAG_5_max_score + U2_GTAG_3_max_score
exons = set([])
for row in csv.reader(open(gencode_bed), delimiter = '\t'):
csv.field_size_limit(1000000000)
qstarts = map (int, row[11].strip(",").split(","))
blocksizes = map(int, row[10].strip(",").split(","))
start = int(row[1])
strand = row[5]
bn = int(row[9])
chr = row[0]
for q1, b in zip(qstarts[1:-1], blocksizes[1:-1]):
estart = start + q1
eend = start + q1 + b
E5 = str(Genome[chr][estart-14:estart+3]).upper()
E3 = str(Genome[chr][eend-3:eend+10]).upper()
if strand == "-":
E5 = str(Genome[chr][eend-3:eend+14].reverse_complement()).upper()
E3 = str(Genome[chr][estart-10:estart+3].reverse_complement()).upper()
U2_score = 0
i = 0
for N in E5:
U2_score += U2_GTAG_3[N][i]
i += 1
i = 0
for N in E3:
U2_score += U2_GTAG_5[N][i]
i += 1
U2_score = percent(U2_score, TOTAL_U2_max_score)
if E5[-5:-3]=="AG" and E3[3:5] == "GT":
exons.add((chr, estart, eend, strand, U2_score))
# if " ".join([chr, estart, eend]) == "chr17 26597935 26598725":
# print
for e in exons:
chr, estart, eend, strand, U2_score = e
conservation_vertebrates = wWigIO.getIntervals(phylop_vertebrates, chr, estart-2, eend+2)
conservation_primates = wWigIO.getIntervals(phylop_primates, chr, estart-2, eend+2)
mean_conservation_vertebrates = 0
mean_conservation_primates = 0
for i in conservation_vertebrates:
mean_conservation_vertebrates += i[2]
try:
mean_conservation_vertebrates = mean_conservation_vertebrates/len(conservation_vertebrates)
except ZeroDivisionError:
pass
for i in conservation_primates:
mean_conservation_primates += i[2]
try:
mean_conservation_primates = mean_conservation_primates/len(conservation_primates)
except ZeroDivisionError:
pass
print chr, estart, eend, strand, U2_score, mean_conservation_vertebrates, mean_conservation_primates