def variant_queryable():
vcf = MultiSampleVCF(vcf_file)
return VariantIntervalQueryable(
vcf, [([
Variant('chr1', 12, 'A', 'T'),
Variant('chr1', 18, 'A', 'C', filter='q10'),
], Interval('chr1', 10, 20)),
([
Variant('chr2', 120, 'AT', 'AAAT'),
], Interval('chr2', 110, 200))])
def test_MultiSampleVCF__regions_from_variants(multi_sample_vcf):
variants = [
Variant('chr1', 4, 'T', 'C'),
Variant('chr1', 25, 'AACG', 'GA'),
Variant('chr1', 55525, 'AACG', 'GA'),
Variant('chr10', 55525, 'AACG', 'GA')
]
regions = multi_sample_vcf._regions_from_variants(variants)
assert set(regions) == set([
Interval('chr1', 3, 25),
Interval('chr1', 55524, 55525),
Interval('chr10', 55524, 55525)
])
def get_variants(self, variants: Iterable[Union[str, Variant]],
regions=None, variant_gap=150) -> List[Variant]:
"""Returns list of variants from vcf file. Lets you use vcf file as dict.
# Arguments:
variants: list of variants
regions: list of regions to seek for variants.
Automatically generated from variants if not given.
strategy: strategy if there is not variant in region.
# Returns
List of variants
"""
variants = [
Variant.from_str(v) if type(v) == str else v
for v in variants
]
regions = regions or self._regions_from_variants(
variants, variant_gap=variant_gap)
variant_map = dict()
for r in regions:
r_variants = self.fetch_variants(r)
for v in r_variants:
variant_map[v] = v
return [variant_map.get(v) for v in variants]
def _variants_from_cyvcf2(self, cy_variant):
# in case deletion is present
ALTs = cy_variant.ALT or ['']
# single REF can have multiple ALT
for alt in ALTs:
v = Variant.from_cyvcf_and_given_alt(cy_variant, alt)
if 'N' in alt or '*' in alt:
logging.warning(
'Undefined variant %s are not supported: Skip' % str(v))
continue
yield v
def test_MultiSampleVCF_get_variant(multi_sample_vcf):
variant = multi_sample_vcf.get_variant("chr1:4:T>C")
assert variant.chrom == 'chr1'
assert variant.pos == 4
assert variant.ref == 'T'
assert variant.alt == 'C'
variant = multi_sample_vcf.get_variant(Variant('chr1', 4, 'T', 'C'))
assert variant.chrom == 'chr1'
assert variant.pos == 4
assert variant.ref == 'T'
assert variant.alt == 'C'
with pytest.raises(KeyError):
multi_sample_vcf.get_variant("chr1:4:A>C")
def _get_sample_variants(self, variants, sample, phase):
"""Given a list of `cyvcf2.Variant`, returns all those present for a
given sample and phase and converts them to
`kipoiseq.dataclasses.Variant`
Args:
variants: List of `cyvcf2.Variant`, Variants of interest
sample: `str`, Sample for which to filter genotypes
phase: `0` or `1`, Phase for which to filter genotypes
Returns:
List of `kipoiseq.dataclasses.Variant`
"""
sample_index = self._sample_indices[sample]
return [
Variant.from_cyvcf(v) for v in variants
if v.genotypes[sample_index][phase]
]
def get_variant(self, variant: Union[Variant, str]) -> Variant:
"""Returns variant from vcf file. Lets you use vcf file as dict.
# Arguments:
variant: variant object or variant id as string.
# Returns
Variant object.
# Example
```python
>>> MultiSampleVCF(vcf_path).get_variant("chr1:4:T:['C']")
```
"""
if type(variant) == str:
variant = Variant.from_str(variant)
variants = self.fetch_variants(
Interval(variant.chrom, variant.pos - 1, variant.pos))
for v in variants:
if v.ref == variant.ref and v.alt == variant.alt:
return v
raise KeyError('Variant %s not found in vcf file.' % str(variant))
import pytest
from conftest import vcf_file, gtf_file, example_intervals_bed
import pyranges
from kipoiseq.dataclasses import Interval, Variant
from kipoiseq.extractors.vcf import MultiSampleVCF
from kipoiseq.extractors.vcf_matching import variants_to_pyranges, \
pyranges_to_intervals, intervals_to_pyranges, BaseVariantMatcher, \
SingleVariantMatcher, MultiVariantsMatcher, VariantFetcher
intervals = [
Interval('chr1', 1, 10, strand='+'),
Interval('chr1', 23, 30, strand='-')
]
variants = [
Variant('chr1', 4, 'T', 'C'),
Variant('chr1', 5, 'A', 'GA'),
Variant('chr1', 25, 'AACG', 'GA')
]
pr = pyranges.PyRanges(chromosomes='chr1',
starts=[1, 23, 5],
ends=[10, 30, 50],
strands=['+', '-', '.'])
class VariantFetcherProxy(VariantFetcher):
def __init__(self, variant_fetcher: VariantFetcher):
self.variant_fetcher = variant_fetcher
def fetch_variants(
def __next__(self):
return Variant.from_cyvcf(super().__next__())
def test_extract(variant_seq_extractor):
variants = [Variant.from_cyvcf(v) for v in VCF(vcf_file)]
interval = Interval('chr1', 2, 9)
seq = variant_seq_extractor.extract(interval, variants, anchor=5)
assert len(seq) == interval.end - interval.start
assert seq == 'CGAACGT'
interval = Interval('chr1', 2, 9, strand='-')
seq = variant_seq_extractor.extract(interval, variants, anchor=5)
assert len(seq) == interval.end - interval.start
assert seq == 'ACGTTCG'
interval = Interval('chr1', 4, 14)
seq = variant_seq_extractor.extract(interval, variants, anchor=7)
assert len(seq) == interval.end - interval.start
assert seq == 'AACGTAACGT'
interval = Interval('chr1', 4, 14)
seq = variant_seq_extractor.extract(interval, variants, anchor=4)
assert len(seq) == interval.end - interval.start
assert seq == 'GAACGTAACG'
interval = Interval('chr1', 2, 5)
seq = variant_seq_extractor.extract(interval, variants, anchor=3)
assert len(seq) == interval.end - interval.start
assert seq == 'GCG'
interval = Interval('chr1', 24, 34)
seq = variant_seq_extractor.extract(interval, variants, anchor=27)
assert len(seq) == interval.end - interval.start
assert seq == 'TGATAACGTA'
interval = Interval('chr1', 25, 35)
seq = variant_seq_extractor.extract(interval, variants, anchor=34)
assert len(seq) == interval.end - interval.start
assert seq == 'TGATAACGTA'
interval = Interval('chr1', 34, 44)
seq = variant_seq_extractor.extract(interval, variants, anchor=37)
assert len(seq) == interval.end - interval.start
assert seq == 'AACGTAACGT'
interval = Interval('chr1', 34, 44)
seq = variant_seq_extractor.extract(interval, variants, anchor=100)
assert len(seq) == interval.end - interval.start
assert seq == 'AACGTAACGT'
interval = Interval('chr1', 5, 11, strand='+')
seq = variant_seq_extractor.extract(interval,
variants,
anchor=10,
fixed_len=False)
assert seq == 'ACGTAA'
interval = Interval('chr1', 0, 3, strand='+')
seq = variant_seq_extractor.extract(interval,
variants,
anchor=10,
fixed_len=False)
assert seq == 'ACG'
interval = Interval('chr1', 0, 3, strand='+')
ref_seq_extractor = FastaStringExtractor(fasta_file, use_strand=True)
seq = VariantSeqExtractor(reference_sequence=ref_seq_extractor).extract(
interval, variants, anchor=10, fixed_len=False)
assert seq == 'ACG'
def fetch_variants(self, interval, sample_id=None):
for v in self(self._region(interval)):
v = Variant.from_cyvcf(v)
if sample_id is None or self.has_variant(v, sample_id):
yield v
def test_variant():
v = Variant("chr1", 10, 'C', 'T')
assert v.start == 9
assert v.chrom == 'chr1'
assert v.pos == 10
assert v.ref == 'C'
assert v.alt == 'T'
assert isinstance(v.info, dict)
assert len(v.info) == 0
assert v.qual == 0
assert v.filter == 'PASS'
v.info['test'] = 10
assert v.info['test'] == 10
assert isinstance(str(v), str)
# make sure the original got unchangd
v2 = v.copy()
v.info['test'] = 20
assert v2.info['test'] == 10
v.__repr__()
# __str__, from_str
assert v == Variant.from_str(str(v))
# hash test
assert isinstance(hash(v), int)
assert hash(v) == hash(Variant.from_str(str(v)))
# fixed arguments
with pytest.raises(AttributeError):
v.chrom = 'asd'
with pytest.raises(AttributeError):
v.pos = 10
with pytest.raises(AttributeError):
v.ref = 'asd'
with pytest.raises(AttributeError):
v.alt = 'asd'
# non-fixed arguments
v.id = 'asd'
v.qual = 10
v.filter = 'asd'
v.source = 2
assert isinstance(Variant("chr1", '10', 'C', 'T').pos, int)
# from cyvcf2
vcf = cyvcf2.VCF('tests/data/test.vcf.gz')
cv = list(vcf)[0]
v2 = Variant.from_cyvcf(cv)
assert isinstance(v2.source, cyvcf2.Variant)