def extract_transposon_fusions(chimeric_data,
transposon_name,
merge_junction_dist=10,
max_spanning_dist=300,
max_junction_dist=10000):
"""Extracts transposon fusions from a STAR chimeric read dataframe."""
# Subset and normalize chimeric reads for transposon.
chimeric_data = chimeric_data.loc[
(chimeric_data['seqname_a'] == transposon_name) ^
(chimeric_data['seqname_b'] == transposon_name)] # yapf: disable
chimeric_data = normalize_chimeric_junctions(chimeric_data,
seqname=transposon_name)
# Extract junction fusions and merge close junctions.
junctions = list(
extract_junction_fusions(chimeric_data,
merge_dist=merge_junction_dist))
# Assign spanning reads to junctions.
junctions, unassigned = assign_spanning_reads(
junctions,
chimeric_data,
max_dist_left=max_spanning_dist,
max_dist_right=max_junction_dist)
# Extract spanning fusions from unused reads.
spanning = extract_spanning_fusions(unassigned, max_dist=max_spanning_dist)
fusions = itertools.chain.from_iterable([junctions, spanning])
# Convert to transposon fusions.
for fusion in fusions:
yield TransposonFusion.from_fusion(fusion, transposon_name)
def _to_fusion_obj(fusion, transposon_name, is_paired):
if fusion.seqname_a == transposon_name:
gen_id, tr_id = 'b', 'a'
gen_dir, tr_dir = 1, -1
else:
gen_id, tr_id = 'a', 'b'
gen_dir, tr_dir = -1, 1
strand_genome = fusion['strand_' + gen_id]
strand_transposon = fusion['strand_' + tr_id]
if is_paired:
support_junction = fusion.supp_spanning_mates
support_spanning = fusion.supp_mates
else:
support_junction = fusion.supp_reads
support_spanning = 0
return TransposonFusion(
seqname=fusion['seqname_' + gen_id],
anchor_genome=fusion['location_' + gen_id],
anchor_transposon=fusion['location_' + tr_id],
flank_genome=fusion['flank_' + gen_id] * strand_genome * gen_dir,
flank_transposon=fusion['flank_' + tr_id] * strand_transposon * tr_dir,
strand_genome=strand_genome,
strand_transposon=strand_transposon,
support_junction=support_junction,
support_spanning=support_spanning,
metadata=frozendict())
def extract_transposon_fusions(chimeric_data,
transposon_name,
merge_junction_dist=10,
max_spanning_dist=300,
max_junction_dist=10000):
"""Extracts transposon fusions from a STAR chimeric read dataframe."""
# Subset and normalize chimeric reads for transposon.
chimeric_data = chimeric_data.loc[
(chimeric_data['seqname_a'] == transposon_name) ^
(chimeric_data['seqname_b'] == transposon_name)] # yapf: disable
chimeric_data = normalize_chimeric_junctions(
chimeric_data, seqname=transposon_name)
# Extract junction fusions and merge close junctions.
junctions = list(
extract_junction_fusions(
chimeric_data, merge_dist=merge_junction_dist))
# Assign spanning reads to junctions.
junctions, unassigned = assign_spanning_reads(
junctions,
chimeric_data,
max_dist_left=max_spanning_dist,
max_dist_right=max_junction_dist)
# Extract spanning fusions from unused reads.
spanning = extract_spanning_fusions(unassigned, max_dist=max_spanning_dist)
fusions = itertools.chain.from_iterable([junctions, spanning])
# Convert to transposon fusions.
for fusion in fusions:
yield TransposonFusion.from_fusion(fusion, transposon_name)
def test_from_fusion(self, fusion):
"""Tests from_fusion method."""
tr_fusion = TransposonFusion.from_fusion(fusion, 'T2onc')
expected = TransposonFusion(seqname='1',
anchor_genome=300,
anchor_transposon=420,
strand_genome=1,
strand_transposon=1,
flank_genome=-52,
flank_transposon=62,
support_junction=2,
support_spanning=1,
metadata=frozendict({}))
assert tr_fusion == expected
assert isinstance(tr_fusion.metadata, frozendict)
def rgag1_fusion():
return TransposonFusion(
seqname='X',
anchor_genome=143093898,
anchor_transposon=1541,
strand_genome=1,
strand_transposon=1,
flank_genome=-78,
flank_transposon=-76,
support_junction=10,
support_spanning=0,
metadata=frozendict({}))
def fusion():
"""Example fusion."""
return TransposonFusion(
seqname='16',
anchor_genome=52141095,
anchor_transposon=1541,
strand_genome=-1,
strand_transposon=1,
flank_genome=-78,
flank_transposon=-76,
support_junction=380,
support_spanning=118,
metadata=frozendict({}))
def test_example(self, chimeric_data):
"""Test simple example."""
fusions = list(star.extract_transposon_fusions(chimeric_data, 'T2onc'))
assert fusions == [
TransposonFusion(
seqname='1',
anchor_genome=300,
anchor_transposon=420,
strand_genome=-1,
strand_transposon=-1,
flank_genome=-62,
flank_transposon=100,
support_junction=4,
support_spanning=3,
metadata=frozendict({}))
]
def tr_fusion():
"""Example transposon fusion."""
return TransposonFusion(seqname='16',
anchor_genome=52141095,
anchor_transposon=1541,
strand_genome=-1,
strand_transposon=1,
flank_genome=-78,
flank_transposon=-76,
support_junction=380,
support_spanning=118,
metadata=frozendict({
'gene_strand': 1,
'gene_name': 'Cblb',
'feature_type': 'SA',
'feature_strand': -1,
'feature_name': 'En2SA'
}))
def test_from_fusion(self, fusion):
"""Tests from_fusion method."""
tr_fusion = TransposonFusion.from_fusion(fusion, 'T2onc')
expected = TransposonFusion(
seqname='1',
anchor_genome=300,
anchor_transposon=420,
strand_genome=1,
strand_transposon=1,
flank_genome=-52,
flank_transposon=62,
support_junction=2,
support_spanning=1,
metadata=frozendict({}))
assert tr_fusion == expected
assert isinstance(tr_fusion.metadata, frozendict)
def test_example_file(self, chimeric_junctions_path):
"""Test data read from example file."""
# Extract fusions.
chimeric_data = star.read_chimeric_junctions(chimeric_junctions_path)
fusions = set(star.extract_transposon_fusions(chimeric_data, 'T2onc'))
# Check number of fusions.
assert len(fusions) == 7
# Check key example.
expected = TransposonFusion(
seqname='16',
anchor_genome=52141095,
anchor_transposon=1541,
strand_genome=-1,
strand_transposon=1,
flank_genome=-78,
flank_transposon=-76,
support_junction=380,
support_spanning=118,
metadata=frozendict({}))
assert expected in fusions