def test_parse_literal_with_contig_map(self, contig_name, expected):
contig_map = {
'chr1': core_pb2.ContigInfo(name='chr1', n_bases=10),
'chr2': core_pb2.ContigInfo(name='chr2', n_bases=5),
}
self.assertEqual(
ranges.parse_literal(contig_name, contig_map=contig_map), expected)
def setUp(self):
self.proto1 = core_pb2.ContigInfo(name='p1',
n_bases=10,
pos_in_fasta=0)
self.proto2 = core_pb2.ContigInfo(name='p2',
n_bases=20,
pos_in_fasta=1)
self.protos = [self.proto1, self.proto2]
def test_from_contigs(self):
contigs = [
core_pb2.ContigInfo(name='chr1', n_bases=10),
core_pb2.ContigInfo(name='chr2', n_bases=5),
]
self.assertCountEqual([
ranges.make_range('chr1', 0, 10),
ranges.make_range('chr2', 0, 5),
], ranges.RangeSet.from_contigs(contigs))
def test_from_regions_not_empty(self):
literals = ['chr1', 'chr2:10-20']
contig_map = {
'chr1': core_pb2.ContigInfo(name='chr1', n_bases=10),
'chr2': core_pb2.ContigInfo(name='chr2', n_bases=100),
}
self.assertItemsEqual(
[ranges.make_range('chr1', 0, 10),
ranges.make_range('chr2', 9, 20)],
ranges.RangeSet.from_regions(literals, contig_map))
def test_contigs_n_bases(self): c1 = core_pb2.ContigInfo(name='c', n_bases=100, pos_in_fasta=0) c2 = core_pb2.ContigInfo(name='a', n_bases=50, pos_in_fasta=1) c3 = core_pb2.ContigInfo(name='b', n_bases=25, pos_in_fasta=2) self.assertEqual(100, ranges.contigs_n_bases([c1])) self.assertEqual(50, ranges.contigs_n_bases([c2])) self.assertEqual(25, ranges.contigs_n_bases([c3])) self.assertEqual(150, ranges.contigs_n_bases([c1, c2])) self.assertEqual(125, ranges.contigs_n_bases([c1, c3])) self.assertEqual(175, ranges.contigs_n_bases([c1, c2, c3]))
def setUp(self):
self.out_fname = test_utils.test_tmpfile('output.vcf')
self.options = core_pb2.VcfWriterOptions(
contigs=[
core_pb2.ContigInfo(name='Chr1', n_bases=50, pos_in_fasta=0),
core_pb2.ContigInfo(name='Chr2', n_bases=25, pos_in_fasta=1),
],
sample_names=['Fido', 'Spot'],
filters=[])
self.writer = vcf_writer.VcfWriter.to_file(self.out_fname, self.options)
self.variant = test_utils.make_variant(
chrom='Chr1', start=10, alleles=['A', 'C'])
self.variant.calls.add(genotype=[0, 0], call_set_name='Fido')
self.variant.calls.add(genotype=[0, 1], call_set_name='Spot')
def test_parse_literal_with_contig_map_and_bad_input_raises_exception(
self, bad_literal):
with self.assertRaises(ValueError):
ranges.parse_literal(
bad_literal,
contig_map={
'chr1': core_pb2.ContigInfo(name='chr1', n_bases=10)
})
def write_variant_to_tempfile(self, variant):
path = test_utils.test_tmpfile('test.vcf')
writer = genomics_io.make_vcf_writer(
outfile=path,
contigs=[core_pb2.ContigInfo(name='20')],
samples=[call.call_set_name for call in variant.calls],
filters=[])
with writer:
writer.write(variant)
return path
def setUp(self):
self.read1 = test_utils.make_read(bases='ACCGT',
chrom='chr1',
start=10,
cigar='5M',
mapq=50,
quals=range(30, 35),
name='read1')
self.read2 = test_utils.make_read(bases='AACCTT',
chrom='chr2',
start=15,
cigar='7M',
mapq=40,
quals=range(20, 26),
name='read2')
self.contigs = [
core_pb2.ContigInfo(name='chr1'),
core_pb2.ContigInfo(name='chr2'),
]
def test_sort_ranges(self):
contigs = [
core_pb2.ContigInfo(name='c', n_bases=100, pos_in_fasta=0),
core_pb2.ContigInfo(name='a', n_bases=76, pos_in_fasta=1),
core_pb2.ContigInfo(name='b', n_bases=121, pos_in_fasta=2),
]
unsorted = ranges.parse_literals(
['a:10', 'c:20', 'b:30', 'b:10-15', 'b:10', 'a:5'])
# Without contigs we sort the contigs by name lexicographically.
self.assertEqual(
ranges.parse_literals(
['a:5', 'a:10', 'b:10', 'b:10-15', 'b:30', 'c:20']),
ranges.sorted_ranges(unsorted))
# With contigs we sort by the position of the contigs themselves.
self.assertEqual(
ranges.parse_literals(
['c:20', 'a:5', 'a:10', 'b:10', 'b:10-15', 'b:30']),
ranges.sorted_ranges(unsorted, contigs))
def _make_contigs(specs):
"""Makes ContigInfo protos from specs.
Args:
specs: A list of 2- or 3-tuples. All tuples should be of the same length.
If 2-element, these should be the name and length in basepairs of each
contig, and their pos_in_fasta will be set to their index in the list. If
the 3-element, the tuple should contain name, length, and pos_in_fasta.
Returns:
A list of ContigInfo protos, one for each spec in specs.
"""
if specs and len(specs[0]) == 3:
return [
core_pb2.ContigInfo(name=name, n_bases=length, pos_in_fasta=i)
for name, length, i in specs
]
else:
return [
core_pb2.ContigInfo(name=name, n_bases=length, pos_in_fasta=i)
for i, (name, length) in enumerate(specs)
]
def test_writing_canned_variants(self):
"""Tests writing all the variants that are 'canned' in our tfrecord file."""
# This file is in the TF record format
tfrecord_file = test_utils.genomics_core_testdata(
'test_samples.vcf.golden.tfrecord')
writer_options = core_pb2.VcfWriterOptions(
contigs=[
core_pb2.ContigInfo(name='chr1', n_bases=248956422),
core_pb2.ContigInfo(name='chr2', n_bases=242193529),
core_pb2.ContigInfo(name='chr3', n_bases=198295559),
core_pb2.ContigInfo(name='chrX', n_bases=156040895)
],
sample_names=['NA12878_18_99'],
filters=[
core_pb2.VcfFilterInfo(id='LowQual'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL95.00to96.00'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL96.00to97.00'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL97.00to99.00'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL99.00to99.50'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL99.50to99.90'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL99.90to99.95'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL99.95to100.00+'),
core_pb2.VcfFilterInfo(id='VQSRTrancheINDEL99.95to100.00'),
core_pb2.VcfFilterInfo(id='VQSRTrancheSNP99.50to99.60'),
core_pb2.VcfFilterInfo(id='VQSRTrancheSNP99.60to99.80'),
core_pb2.VcfFilterInfo(id='VQSRTrancheSNP99.80to99.90'),
core_pb2.VcfFilterInfo(id='VQSRTrancheSNP99.90to99.95'),
core_pb2.VcfFilterInfo(id='VQSRTrancheSNP99.95to100.00+'),
core_pb2.VcfFilterInfo(id='VQSRTrancheSNP99.95to100.00'),
])
variant_records = list(
io_utils.read_tfrecords(tfrecord_file, proto=variants_pb2.Variant))
out_fname = test_utils.test_tmpfile('output.vcf')
with vcf_writer.VcfWriter.to_file(out_fname, writer_options) as writer:
for record in variant_records[:5]:
writer.write(record)
# Check: are the variants written as expected?
# pylint: disable=line-too-long
expected_vcf_content = [
'##fileformat=VCFv4.2\n',
'##FILTER=<ID=PASS,Description="All filters passed">\n',
'##FILTER=<ID=LowQual,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL95.00to96.00,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL96.00to97.00,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL97.00to99.00,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL99.00to99.50,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL99.50to99.90,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL99.90to99.95,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL99.95to100.00+,Description="">\n',
'##FILTER=<ID=VQSRTrancheINDEL99.95to100.00,Description="">\n',
'##FILTER=<ID=VQSRTrancheSNP99.50to99.60,Description="">\n',
'##FILTER=<ID=VQSRTrancheSNP99.60to99.80,Description="">\n',
'##FILTER=<ID=VQSRTrancheSNP99.80to99.90,Description="">\n',
'##FILTER=<ID=VQSRTrancheSNP99.90to99.95,Description="">\n',
'##FILTER=<ID=VQSRTrancheSNP99.95to100.00+,Description="">\n',
'##FILTER=<ID=VQSRTrancheSNP99.95to100.00,Description="">\n',
'##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">\n',
'##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype Quality">\n',
'##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read depth of all '
'passing filters reads.">\n',
'##FORMAT=<ID=AD,Number=R,Type=Integer,Description="Read depth of all '
'passing filters reads for each allele.">\n',
'##FORMAT=<ID=VAF,Number=A,Type=Float,Description=\"Variant allele '
'fractions.">\n',
'##FORMAT=<ID=GL,Number=G,Type=Float,Description="Genotype '
'likelihoods, log10 encoded">\n',
'##FORMAT=<ID=PL,Number=G,Type=Integer,Description="Genotype '
'likelihoods, Phred encoded">\n',
'##INFO=<ID=END,Number=1,Type=Integer,Description="Stop position of '
'the interval">\n', '##contig=<ID=chr1,length=248956422>\n',
'##contig=<ID=chr2,length=242193529>\n',
'##contig=<ID=chr3,length=198295559>\n',
'##contig=<ID=chrX,length=156040895>\n',
'#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\tNA12878_18_99\n',
'chr1\t13613\t.\tT\tA\t39.88\tVQSRTrancheSNP99.90to99.95\t.\tGT:GQ:DP:AD:PL\t0/1:16:4:1,3:68,0,16\n',
'chr1\t13813\t.\tT\tG\t90.28\tPASS\t.\tGT:GQ:DP:AD:PL\t1/1:9:3:0,3:118,9,0\n',
'chr1\t13838\trs28428499\tC\tT\t62.74\tPASS\t.\tGT:GQ:DP:AD:PL\t1/1:6:2:0,2:90,6,0\n',
'chr1\t14397\trs756427959\tCTGT\tC\t37.73\tPASS\t.\tGT:GQ:DP:AD:PL\t0/1:75:5:3,2:75,0,152\n',
'chr1\t14522\t.\tG\tA\t49.77\tVQSRTrancheSNP99.60to99.80\t.\tGT:GQ:DP:AD:PL\t0/1:78:10:6,4:78,0,118\n'
]
# pylint: enable=line-too-long
with tf.gfile.GFile(out_fname, 'r') as f:
self.assertEqual(f.readlines(), expected_vcf_content)
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
from absl.testing import absltest
from deepvariant.core import ranges
from deepvariant.core import test_utils
from deepvariant.core.protos import core_pb2
from deepvariant.core.python import vcf_reader
expected_sites_contigs = [
core_pb2.ContigInfo(name='chr1', pos_in_fasta=0, n_bases=248956422),
core_pb2.ContigInfo(name='chr2', pos_in_fasta=1, n_bases=242193529),
core_pb2.ContigInfo(name='chr3', pos_in_fasta=2, n_bases=198295559),
core_pb2.ContigInfo(name='chr4', pos_in_fasta=3, n_bases=190214555),
core_pb2.ContigInfo(name='chr5', pos_in_fasta=4, n_bases=181538259),
core_pb2.ContigInfo(name='chr6', pos_in_fasta=5, n_bases=170805979),
core_pb2.ContigInfo(name='chr7', pos_in_fasta=6, n_bases=159345973),
core_pb2.ContigInfo(name='chr8', pos_in_fasta=7, n_bases=145138636),
core_pb2.ContigInfo(name='chr9', pos_in_fasta=8, n_bases=138394717),
core_pb2.ContigInfo(name='chr10', pos_in_fasta=9, n_bases=133797422),
core_pb2.ContigInfo(name='chr11', pos_in_fasta=10, n_bases=135086622),
core_pb2.ContigInfo(name='chr12', pos_in_fasta=11, n_bases=133275309),
core_pb2.ContigInfo(name='chr13', pos_in_fasta=12, n_bases=114364328),
core_pb2.ContigInfo(name='chr14', pos_in_fasta=13, n_bases=107043718),
core_pb2.ContigInfo(name='chr15', pos_in_fasta=14, n_bases=101991189),
core_pb2.ContigInfo(name='chr16', pos_in_fasta=15, n_bases=90338345),
from deepvariant import postprocess_variants
from deepvariant import test_utils
from deepvariant.core import genomics_math
from deepvariant.core import io_utils
from deepvariant.core import variantutils
from deepvariant.core.protos import core_pb2
from deepvariant.protos import deepvariant_pb2
from deepvariant.testing import flagsaver
FLAGS = flags.FLAGS
_DEFAULT_SAMPLE_NAME = 'NA12878'
# Test contigs for gVCF merging code.
_CONTIGS = [
core_pb2.ContigInfo(name='1', n_bases=100),
core_pb2.ContigInfo(name='2', n_bases=200),
core_pb2.ContigInfo(name='10', n_bases=300),
]
def setUpModule():
test_utils.init()
def _create_variant(ref_name, start, ref_base, alt_bases, qual, filter_field,
genotype, gq, likelihoods):
"""Creates a Variant record for testing.
Args:
ref_name: reference name for this variant
def test_sam_contigs(self):
reader = sam_reader.SamReader.from_file(self.bam, self.options)
with reader:
self.assertEqual([
core_pb2.ContigInfo(name='chrM', pos_in_fasta=0,
n_bases=16571),
core_pb2.ContigInfo(
name='chr1', pos_in_fasta=1, n_bases=249250621),
core_pb2.ContigInfo(
name='chr2', pos_in_fasta=2, n_bases=243199373),
core_pb2.ContigInfo(
name='chr3', pos_in_fasta=3, n_bases=198022430),
core_pb2.ContigInfo(
name='chr4', pos_in_fasta=4, n_bases=191154276),
core_pb2.ContigInfo(
name='chr5', pos_in_fasta=5, n_bases=180915260),
core_pb2.ContigInfo(
name='chr6', pos_in_fasta=6, n_bases=171115067),
core_pb2.ContigInfo(
name='chr7', pos_in_fasta=7, n_bases=159138663),
core_pb2.ContigInfo(
name='chr8', pos_in_fasta=8, n_bases=146364022),
core_pb2.ContigInfo(
name='chr9', pos_in_fasta=9, n_bases=141213431),
core_pb2.ContigInfo(
name='chr10', pos_in_fasta=10, n_bases=135534747),
core_pb2.ContigInfo(
name='chr11', pos_in_fasta=11, n_bases=135006516),
core_pb2.ContigInfo(
name='chr12', pos_in_fasta=12, n_bases=133851895),
core_pb2.ContigInfo(
name='chr13', pos_in_fasta=13, n_bases=115169878),
core_pb2.ContigInfo(
name='chr14', pos_in_fasta=14, n_bases=107349540),
core_pb2.ContigInfo(
name='chr15', pos_in_fasta=15, n_bases=102531392),
core_pb2.ContigInfo(
name='chr16', pos_in_fasta=16, n_bases=90354753),
core_pb2.ContigInfo(
name='chr17', pos_in_fasta=17, n_bases=81195210),
core_pb2.ContigInfo(
name='chr18', pos_in_fasta=18, n_bases=78077248),
core_pb2.ContigInfo(
name='chr19', pos_in_fasta=19, n_bases=59128983),
core_pb2.ContigInfo(
name='chr20', pos_in_fasta=20, n_bases=63025520),
core_pb2.ContigInfo(
name='chr21', pos_in_fasta=21, n_bases=48129895),
core_pb2.ContigInfo(
name='chr22', pos_in_fasta=22, n_bases=51304566),
core_pb2.ContigInfo(
name='chrX', pos_in_fasta=23, n_bases=155270560),
core_pb2.ContigInfo(
name='chrY', pos_in_fasta=24, n_bases=59373566),
], reader.contigs)
def write_test_protos(self, filename):
protos = [core_pb2.ContigInfo(name=str(i)) for i in range(10)]
path = test_utils.test_tmpfile(filename)
io.write_tfrecords(protos, path)
return protos, path
