def g_duplex(self):
"""Generator of duplex by gaps"""
segs = self.sgc.segments
for i in range(1, len(segs)):
lftsgc = SegmentChain(*segs[:i], ID="%dLeft" % i)
rgtsgc = SegmentChain(*segs[i:], ID="%dRight" % i)
yield Duplex(lftsgc, rgtsgc, self)
def as_bed(self):
sc = SegmentChain(self.don, self.acc)
ts = sc.spanning_segment.start
te = sc.spanning_segment.end
b = sc.as_bed(thickstart=ts, thickend=te).strip("\n").split("\t")
b.extend(map(str, [self.nevi, self.ntot]))
return b
def get_jnc(self, idx):
assert 1 <= idx <= self.numjnc
segs = self.sgc.segments
s = SegmentChain(*segs[:idx]).spanning_segment.end
e = SegmentChain(*segs[idx:]).spanning_segment.start
gs = GenomicSegment(self.chrom, s, e, self.strand)
return Jnc(gs, self, idx)
def __init__(self, r):
assert isinstance(r, pysam.AlignedSegment)
self.r = r
self.chrom = r.reference_name
self.strand = '-' if self.r.is_reverse else '+'
self.blocks = self.r.blocks
self.tags = {tag: val for tag, val in self.r.tags}
self.sgc = SegmentChain()
for (start, end) in self.r.blocks:
self.sgc.add_segments(
GenomicSegment(self.chrom, start, end, self.strand))
class JncGroup:
"""Junction Group"""
def __init__(self, js):
self.jsli = [js]
self.corejs = js
self.sc = SegmentChain(js.don, js.acc)
self.name = None
def distance2(self, js):
return max(self.corejs.dist2don(js), self.corejs.dist2acc(js))
def add(self, js):
self.jsli.append(js)
def merge(self):
for i in range(1, len(self.jsli)):
js = self.jsli[i]
self.sc.add_segments(js.don)
self.sc.add_segments(js.acc)
def get_ntot(self):
return sum(js.ntot for js in self.jsli)
def get_nevi(self):
return sum(js.nevi for js in self.jsli)
def set_name(self, name):
self.name = name
def as_bed(self):
if self.name is not None:
self.sc.attr['ID'] = self.name
if self.sc.strand == "+":
color = "255,0,0"
else:
color = "0,0,255"
self.sc.attr['color'] = color
b = self.sc.as_bed().strip("\n").split("\t")
b.extend(map(str, [self.get_nevi(), self.get_ntot()]))
return b
def __str__(self):
return "\t".join(
map(str,
[self.corejs, self.get_ntot(),
self.get_nevi()]))
def jncsites(self):
jncs = []
for i in range(len(self.jsli)):
jncs.append([self.name, str(i + 1)] + self.jsli[i].as_bed())
i += 1
return jncs
class ParisRead:
"""Read from PARIS-seq"""
def __init__(self, r):
assert isinstance(r, pysam.AlignedSegment)
self.r = r
self.chrom = r.reference_name
self.strand = '-' if self.r.is_reverse else '+'
self.blocks = self.r.blocks
self.tags = {tag: val for tag, val in self.r.tags}
self.sgc = SegmentChain()
for (start, end) in self.r.blocks:
self.sgc.add_segments(
GenomicSegment(self.chrom, start, end, self.strand))
def __str__(self):
return "%s\t%s" % (self.r.qname, self.sgc)
def g_duplex(self):
"""Generator of duplex by gaps"""
segs = self.sgc.segments
for i in range(1, len(segs)):
lftsgc = SegmentChain(*segs[:i], ID="%dLeft" % i)
rgtsgc = SegmentChain(*segs[i:], ID="%dRight" % i)
yield Duplex(lftsgc, rgtsgc, self)
def sgc3():
return SegmentChain(GenomicSegment("chr1", 350, 450, "+"))
def sgc2():
exon1 = GenomicSegment("chr1", 1150, 1200, "+")
exon2 = GenomicSegment("chr1", 1300, 1450, "+")
return SegmentChain(exon1, exon2, ID="SGC2", alias="SGC2")
def sgc1():
exon1 = GenomicSegment("chr1", 100, 200, "+")
exon2 = GenomicSegment("chr1", 300, 400, "+")
return SegmentChain(exon1, exon2, ID="SGC1", alias="SGC1")
def sgc(self):
sgc = SegmentChain()
for (start, end) in self.blocks:
sgc.add_segments(
GenomicSegment(self.chrom, start, end, self.strand))
return sgc
def __init__(self, js):
self.jsli = [js]
self.corejs = js
self.sc = SegmentChain(js.don, js.acc)
self.name = None
def rgt(self):
return SegmentChain(*self.segread.sgc.segments[self.idx:])
def lft(self):
return SegmentChain(*self.segread.sgc.segments[:self.idx])