def insertions():
"""Example insertions in ENSMUSG00000051951 (5 exons)."""
return [
# After third exon.
Insertion(
id='1',
seqname='1',
position=3207327,
strand=1,
support_junction=1,
support_spanning=1,
support=2,
metadata=frozendict({
'gene_id': 'ENSMUSG00000051951',
'sample': 'S1'})),
# After first exon.
Insertion(
id='2',
seqname='1',
position=3214491,
strand=-1,
support_junction=1,
support_spanning=1,
support=2,
metadata=frozendict({
'gene_id': 'ENSMUSG00000051951',
'sample': 'S2'}))
] # yapf: disable
def insertions():
"""Example insertions in ENSMUSG00000051951 (5 exons)."""
return [
# After third exon.
Insertion(
id='1',
seqname='1',
position=3207327,
strand=1,
support_junction=1,
support_spanning=1,
support=2,
metadata=frozendict({
'gene_id': 'ENSMUSG00000051951',
'sample': 'S1'})),
# After first exon.
Insertion(
id='2',
seqname='1',
position=3214491,
strand=-1,
support_junction=1,
support_spanning=1,
support=2,
metadata=frozendict({
'gene_id': 'ENSMUSG00000051951',
'sample': 'S2'}))
] # yapf: disable
def insertions():
"""Example insertion set."""
return [
_insertion(id='1', seqname='1', position=9, strand=1,
metadata=frozendict({'gene_id': 'gene_a', 'sample': 'S1'})),
_insertion(id='2', seqname='1', position=15, strand=-1,
metadata=frozendict({'gene_id': 'gene_b', 'sample': 'S2'}))
] # yapf: disable
def insertions():
"""Example insertion set."""
return [
_insertion(id='1', seqname='1', position=9, strand=1,
metadata=frozendict({'gene_id': 'gene_a', 'sample': 'S1'})),
_insertion(id='2', seqname='1', position=15, strand=-1,
metadata=frozendict({'gene_id': 'gene_b', 'sample': 'S2'}))
] # yapf: disable
def annotate_fusions_for_genes(fusions, reference):
# type: (Iterable[Fusion], TranscriptReference) -> Iterable[Fusion]
"""Annotates fusions with genes overlapped by the genomic fusion site.
Parameters
----------
fusions : iterable[TransposonFusion]
Fusions to annotate.
gtf_path : pathlib.Path
Path to (indexed) gtf file, containing gene exon features.
Yields
------
Fusion
Fusions, annotated with overlapped genes.
"""
for fusion in fusions:
genes = reference.overlap_genes(fusion.genome_region)
if len(genes) > 0:
for gene in genes:
gene_meta = {
'gene_name': gene.name,
'gene_strand': gene.strand,
'gene_id': gene.id
}
merged_meta = toolz.merge(fusion.metadata, gene_meta)
yield fusion._replace(metadata=frozendict(merged_meta))
else:
yield fusion
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 from_fusion(cls, fusion, transposon_name, metadata=None):
"""Converts the fusion to a transposon fusion object."""
if (fusion.seqname_a == transposon_name
and fusion.seqname_b == transposon_name):
raise ValueError('Fusion does not involve genomic sequence')
elif (fusion.seqname_a != transposon_name
and fusion.seqname_b != transposon_name):
raise ValueError('Fusion does not involve transposon')
elif fusion.seqname_a == transposon_name:
tr_key, gen_key = 'a', 'b'
tr_flank = fusion.flank_a * -fusion.strand_a
gen_flank = fusion.flank_b * fusion.strand_b
else:
tr_key, gen_key = 'b', 'a'
tr_flank = fusion.flank_b * fusion.strand_b
gen_flank = fusion.flank_a * -fusion.strand_a
return TransposonFusion(
seqname=getattr(fusion, 'seqname_' + gen_key),
anchor_genome=getattr(fusion, 'location_' + gen_key),
anchor_transposon=getattr(fusion, 'location_' + tr_key),
strand_genome=getattr(fusion, 'strand_' + gen_key),
strand_transposon=getattr(fusion, 'strand_' + tr_key),
flank_genome=gen_flank,
flank_transposon=tr_flank,
support_junction=fusion.support_junction,
support_spanning=fusion.support_spanning,
metadata=frozendict(metadata or {}))
def annotate_fusions_for_genes(fusions, reference):
# type: (Iterable[Fusion], TranscriptReference) -> Iterable[Fusion]
"""Annotates fusions with genes overlapped by the genomic fusion site.
Parameters
----------
fusions : iterable[TransposonFusion]
Fusions to annotate.
gtf_path : pathlib.Path
Path to (indexed) gtf file, containing gene exon features.
Yields
------
Fusion
Fusions, annotated with overlapped genes.
"""
for fusion in fusions:
genes = reference.overlap_genes(fusion.genome_region)
if len(genes) > 0:
for gene in genes:
gene_meta = {
'gene_name': gene.name,
'gene_strand': gene.strand,
'gene_id': gene.id
}
merged_meta = toolz.merge(fusion.metadata, gene_meta)
yield fusion._replace(metadata=frozendict(merged_meta))
else:
yield fusion
def from_fusion(cls, fusion, transposon_name, metadata=None):
"""Converts the fusion to a transposon fusion object."""
if (fusion.seqname_a == transposon_name and
fusion.seqname_b == transposon_name):
raise ValueError('Fusion does not involve genomic sequence')
elif (fusion.seqname_a != transposon_name and
fusion.seqname_b != transposon_name):
raise ValueError('Fusion does not involve transposon')
elif fusion.seqname_a == transposon_name:
tr_key, gen_key = 'a', 'b'
tr_flank = fusion.flank_a * -fusion.strand_a
gen_flank = fusion.flank_b * fusion.strand_b
else:
tr_key, gen_key = 'b', 'a'
tr_flank = fusion.flank_b * fusion.strand_b
gen_flank = fusion.flank_a * -fusion.strand_a
return TransposonFusion(
seqname=getattr(fusion, 'seqname_' + gen_key),
anchor_genome=getattr(fusion, 'location_' + gen_key),
anchor_transposon=getattr(fusion, 'location_' + tr_key),
strand_genome=getattr(fusion, 'strand_' + gen_key),
strand_transposon=getattr(fusion, 'strand_' + tr_key),
flank_genome=gen_flank,
flank_transposon=tr_flank,
support_junction=fusion.support_junction,
support_spanning=fusion.support_spanning,
metadata=frozendict(metadata or {}))
def _to_obj(cls, record):
record_dict = record._asdict()
metadata_fields = [
k for k in record_dict.keys() if k not in set(cls._get_columns())
]
metadata = {k: record_dict.pop(k) for k in metadata_fields}
metadata.pop('Index', None)
return cls(metadata=frozendict(metadata), **record_dict)
def _to_obj(cls, record):
record_dict = record._asdict()
metadata_fields = [
k for k in record_dict.keys() if k not in set(cls._get_columns())
]
metadata = {k: record_dict.pop(k) for k in metadata_fields}
metadata.pop('Index', None)
return cls(metadata=frozendict(metadata), **record_dict)
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 from_transposon_fusion(cls, fusion, id_=None, drop_metadata=None):
"""Converts (annotated) transposon fusion to an insertion.
Requires
Parameters
----------
id : str
ID to use for the insertion.
Returns
-------
Insertion
Insertion object derived from the transposon fusion.
"""
if drop_metadata is None:
drop_metadata = {'strand_genome', 'strand_transposon'}
# Fusion should have annotated transposon feature.
if 'feature_name' not in fusion.metadata:
raise ValueError(
'Fusion does not have an annotated transposon feature')
strand = fusion.strand_genome * fusion.strand_transposon
if 'gene_strand' in fusion.metadata:
gene_strand = fusion.metadata['gene_strand']
orientation = 'sense' if strand == gene_strand else 'antisense'
else:
orientation = None
ins_metadata = toolz.keyfilter(lambda k: k not in drop_metadata,
fusion.metadata)
ins_metadata['transposon_anchor'] = fusion.anchor_transposon
if orientation is not None:
ins_metadata['orientation'] = orientation
return Insertion(
id=id_,
seqname=fusion.seqname,
position=fusion.anchor_genome,
strand=strand,
support_junction=fusion.support_junction,
support_spanning=fusion.support_spanning,
support=fusion.support,
metadata=frozendict(ins_metadata))
def from_transposon_fusion(cls, fusion, id_=None, drop_metadata=None):
"""Converts (annotated) transposon fusion to an insertion.
Requires
Parameters
----------
id : str
ID to use for the insertion.
Returns
-------
Insertion
Insertion object derived from the transposon fusion.
"""
if drop_metadata is None:
drop_metadata = {'strand_genome', 'strand_transposon'}
# Fusion should have annotated transposon feature.
if 'feature_name' not in fusion.metadata:
raise ValueError(
'Fusion does not have an annotated transposon feature')
strand = fusion.strand_genome * fusion.strand_transposon
if 'gene_strand' in fusion.metadata:
gene_strand = fusion.metadata['gene_strand']
orientation = 'sense' if strand == gene_strand else 'antisense'
else:
orientation = None
ins_metadata = toolz.keyfilter(lambda k: k not in drop_metadata,
fusion.metadata)
ins_metadata['transposon_anchor'] = fusion.anchor_transposon
if orientation is not None:
ins_metadata['orientation'] = orientation
return Insertion(id=id_,
seqname=fusion.seqname,
position=fusion.anchor_genome,
strand=strand,
support_junction=fusion.support_junction,
support_spanning=fusion.support_spanning,
support=fusion.support,
metadata=frozendict(ins_metadata))
def annotate_ffpm(fusions, fastq_path):
# type: (Iterable[Fusion], pathlib.Path) -> Iterable[Fusion]
"""Annotates fusions with FFPM (Fusion Fragments Per Million) score."""
# Calculate normalization factor.
n_reads = count_lines(fastq_path) // 4
norm_factor = (1.0 / n_reads) * 1e6
for fusion in fusions:
ffpm_meta = {
'ffpm_junction': fusion.support_junction * norm_factor,
'ffpm_spanning': fusion.support_spanning * norm_factor,
'ffpm': fusion.support * norm_factor
}
merged_meta = toolz.merge(fusion.metadata, ffpm_meta)
yield fusion._replace(metadata=frozendict(merged_meta))
def annotate_ffpm(fusions, fastq_path):
# type: (Iterable[Fusion], pathlib.Path) -> Iterable[Fusion]
"""Annotates fusions with FFPM (Fusion Fragments Per Million) score."""
# Calculate normalization factor.
n_reads = count_lines(fastq_path) // 4
norm_factor = (1.0 / n_reads) * 1e6
for fusion in fusions:
ffpm_meta = {
'ffpm_junction': fusion.support_junction * norm_factor,
'ffpm_spanning': fusion.support_spanning * norm_factor,
'ffpm': fusion.support * norm_factor
}
merged_meta = toolz.merge(fusion.metadata, ffpm_meta)
yield fusion._replace(metadata=frozendict(merged_meta))
def _insertion(id,
seqname,
position,
strand,
support_junction=1,
support_spanning=1,
metadata=None):
"""Helper function for building an Insertion instance."""
return Insertion(id=id,
seqname=seqname,
position=position,
strand=strand,
support_junction=support_junction,
support_spanning=support_spanning,
support=support_junction + support_spanning,
metadata=frozendict(metadata or {}))
def _insertion(id,
seqname,
position,
strand,
support_junction=1,
support_spanning=1,
metadata=None):
"""Helper function for building an Insertion instance."""
return Insertion(
id=id,
seqname=seqname,
position=position,
strand=strand,
support_junction=support_junction,
support_spanning=support_spanning,
support=support_junction + support_spanning,
metadata=frozendict(metadata or {}))
def insertion():
return Insertion(
id='INS_1',
seqname='16',
position=52141095,
strand=-1,
support_junction=380,
support_spanning=118,
support=498,
metadata=frozendict({
'feature_type': 'SA',
'feature_name': 'En2SA',
'gene_strand': 1,
'orientation': 'antisense',
'feature_strand': -1,
'gene_name': 'Cblb',
'transposon_anchor': 1541
}))
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 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 annotate_fusions_for_transposon(fusions, feature_path):
# type: (Iterable[Fusion], pathlib.Path) -> Iterable[Fusion]
"""Annotates fusions with transposon features overlapped by the fusion.
Parameters
----------
fusions : iterable[TransposonFusion]
Fusions to annotate.
feature_path : str or pathlib.Path
Path to TSV file containing transposon features.
Yields
------
Fusion
Fusions, annotated with transposon features.
"""
def _lookup_tree(tree, region):
start, end = region
overlap = tree[start:end]
return [tup[2] for tup in overlap]
# Build the feature tree.
features = TransposonFeature.from_csv(feature_path, sep='\t')
tree = IntervalTree.from_tuples((feat.start, feat.end, feat)
for feat in features)
for fusion in fusions:
overlap = _lookup_tree(tree, fusion.transposon_region)
if len(overlap) > 0:
for feature in overlap:
new_meta = {
'feature_name': feature.name,
'feature_type': feature.type,
'feature_strand': feature.strand
}
merged_meta = toolz.merge(fusion.metadata, new_meta)
yield fusion._replace(metadata=frozendict(merged_meta))
else:
yield fusion
def annotate_fusions_for_transposon(fusions, feature_path):
# type: (Iterable[Fusion], pathlib.Path) -> Iterable[Fusion]
"""Annotates fusions with transposon features overlapped by the fusion.
Parameters
----------
fusions : iterable[TransposonFusion]
Fusions to annotate.
feature_path : str or pathlib.Path
Path to TSV file containing transposon features.
Yields
------
Fusion
Fusions, annotated with transposon features.
"""
def _lookup_tree(tree, region):
start, end = region
overlap = tree[start:end]
return [tup[2] for tup in overlap]
# Build the feature tree.
features = TransposonFeature.from_csv(feature_path, sep='\t')
tree = IntervalTree.from_tuples(
(feat.start, feat.end, feat) for feat in features)
for fusion in fusions:
overlap = _lookup_tree(tree, fusion.transposon_region)
if len(overlap) > 0:
for feature in overlap:
new_meta = {
'feature_name': feature.name,
'feature_type': feature.type,
'feature_strand': feature.strand
}
merged_meta = toolz.merge(fusion.metadata, new_meta)
yield fusion._replace(metadata=frozendict(merged_meta))
else:
yield fusion
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