def getIntSig(f, records, minPts, discut):
"""
@param:discut, distance cutoff determined for self-ligation pets.
"""
print "Starting estimate significance for interactions in %s" % f
model, N = getGenomeCoverage(f, discut)
print "Genomic coverage model built from %s" % f
if N == 0:
print "No cis-PETs parsed as requiring distance cutoff >%s from %s" % (
discut, f)
return None
ds = {}
i = 0
for r in records:
chrom = r[0]
key = "%s-%s-%s" % (r[0], r[3], i)
iva = [r[1] - 1, r[2] + 1]
ivb = [r[4] - 1, r[5] + 1]
#filter loops
distance = abs(sum(ivb) / 2.0 - sum(iva) / 2.0)
if distance < discut:
continue
ra, rb, rab = getPETsforRegions(iva, ivb, model)
#filter clusters contain many self-ligation PETs within distance cutoff
if rab < minPts:
continue
i += 1
if i % 100 == 0:
cFlush("%s interaction p-values estimated for %s" % (i, f))
ra, rb, rab, es, fdr, hyp, chyp, pop, nbp = getMultiplePsFdr(
iva, ivb, model, N)
#this part should be furthur modified, as for most ideable data, there are no noise, so the es should be inf, however, not possible
if es == "None":
continue
ds[key] = {
"distance": distance,
"ra": ra,
"rb": rb,
"rab": rab,
"ES": es,
"FDR": fdr,
"hypergeometric_p-value": hyp,
"hypergeometric_local_FDR": chyp,
"poisson_p-value": pop,
"binomal_p-value": nbp,
"iva": "%s:%s-%s" % (chrom, iva[0], iva[1]),
"ivb": "%s:%s-%s" % (chrom, ivb[0], ivb[1])
}
#memory usage
del model
gc.collect()
print
if len(ds.keys()) == 0:
return None
ds = pd.DataFrame(ds).T
ds["poisson_p-value_corrected"] = getBonPvalues(ds["poisson_p-value"])
ds["binomal_p-value_corrected"] = getBonPvalues(ds["binomal_p-value"])
ds["hypergeometric_p-value_corrected"] = getBonPvalues(
ds["hypergeometric_p-value"])
return ds
def parseRawBedpe2(fs, fout, cs, cut, logger):
"""
Get the cis-PETs, organized by chromosomes. Input could be mixed PETs in bedpe.gz or bedpe. Also change read id to numbers to minize memory usage.
@param fs: bedpe files of replicates, could be .bedpe or .bedpe.gz
@param fout: output prefix, the name for directory
@param cs: chroms that wanted, list like ["chr1","chr2"]
"""
#chroms data
chroms = {}
#cis files
cfs = []
#distance between PETs mapped to different strands
ds = []
i, j, = 0, 0
for f in fs:
r = "Parsing PETs from %s, requiring initial distance cutoff > %s" % (
f, cut)
logger.info(r)
if f.endswith(".gz"):
of = gzip.open(f, "rb")
else:
of = open(f)
for line in of:
i += 1
if i % 100000 == 0:
cFlush("%s PETs processed from %s" % (i, f))
line = line.split("\n")[0].split("\t")
if "*" in line and "-1" in line:
continue
if len(line) < 6:
continue
try:
pet = PET(line)
except:
continue
#cis reads
if pet.chromA != pet.chromB:
continue
#filtering unwanted PETs in chroms
if len(cs) > 0 and (not (pet.chromA in cs and pet.chromB in cs)):
continue
#filtering too close PETs
if cut > 0 and pet.distance < cut:
continue
if pet.chromA not in chroms:
cf = os.path.join(fout,
"%s-%s" % (pet.chromA, pet.chromB) + ".txt")
chroms[pet.chromA] = {"f": open(cf, "w"), "c": 0}
cfs.append(cf)
nline = [chroms[pet.chromA]["c"], pet.cA, pet.cB]
chroms[pet.chromA]["f"].write("\t".join(map(str, nline)) + "\n")
chroms[pet.chromA]["c"] += 1
j += 1
print()
del(chroms)
r = "Totaly %s PETs from %s, in which %s cis PETs" % (i, ",".join(fs), j)
logger.info(r)
return cfs
def estSigOneChr(rs, jdf, pre, dis=0, win=5):
"""
Estimating the significances for the loops in one chromosome.
"""
#all variables with suffix t is treatment, with suffix c in control
logger.info("Building genomic coverage model for %s" % jdf)
model, N = getGenomeCoverage(jdf, dis)
ds = {}
i = 0
for key, r in rs.items():
i += 1
if i % 100 == 0:
report = "Estimating %s loops for %s" % (i, pre)
cFlush(report)
chrom = r[0]
iva = [r[1], r[2]]
ivb = [r[4], r[5]]
ra, rb, rab = getPETsforRegions(iva, ivb, model)
ivas, ivbs = getNearbyPairRegions(iva, ivb, win=win)
mrab = getPermutatedBg(ivas, ivbs, model)
if mrab > 0:
es = rab / mrab
else:
es = 100
pop = max([1e-300, poisson.sf(rab - 1.0, mrab)])
niva = "%s:%s-%s" % (chrom, iva[0], iva[1])
nivb = "%s:%s-%s" % (chrom, ivb[0], ivb[1])
ds[key] = {
"iva": niva,
"ivb": nivb,
"ra": ra,
"rb": rb,
"rab": rab,
"ES": es,
"poisson_p-value": pop,
}
if len(ds) == 0:
return None
ds = pd.DataFrame(ds).T
return ds