def test_generate_tfrecords(self):
examples_out = os.path.join(absltest.get_default_test_tmpdir(),
'examples_output')
train_test_val_split = [0.7, 0.2, 0.1]
ngs_read_length = ngs_errors.generate_tfrecord_datasets(
train_test_val_split,
ref_path=test_utils.genomics_core_testdata(
'ucsc.hg19.chr20.unittest.fasta.gz'),
vcf_path=test_utils.genomics_core_testdata(
'test_nist.b37_chr20_100kbp_at_10mb.vcf.gz'),
bam_path=test_utils.genomics_core_testdata(
'NA12878_S1.chr20.10_10p1mb.bam'),
out_dir=examples_out,
max_reads=100)
actual_examples = self._read_examples(train_test_val_split,
examples_out)
golden_examples = self._read_examples(
train_test_val_split,
test_utils.genomics_core_testdata('golden.examples.ngs_errors'))
self.assertEqual(len(actual_examples), len(golden_examples))
matched_examples = []
for expected in golden_examples:
for actual in actual_examples:
if all(actual.features.feature[key] ==
expected.features.feature[key]
for key in expected.features.feature.keys()):
matched_examples.append(expected)
self.assertEqual(golden_examples, matched_examples)
def test_wrap(self, fasta_filename):
chr_names = ['chrM', 'chr1', 'chr2']
chr_lengths = [100, 76, 121]
fasta = test_utils.genomics_core_testdata(fasta_filename)
fai = test_utils.genomics_core_testdata(fasta_filename + '.fai')
with reference_fai.GenomeReferenceFai.from_file(fasta, fai) as ref:
self.assertEqual(ref.n_contigs, 3)
self.assertIn(fasta, ref.fasta_path)
self.assertIn('GenomeReference backed by htslib FAI index',
str(ref))
self.assertEqual(ref.contig_names, chr_names)
self.assertEqual(ref.n_bp, sum(chr_lengths))
self.assertEqual(ref.bases(ranges.make_range('chrM', 1, 10)),
'ATCACAGGT')
self.assertTrue(
ref.is_valid_interval(ranges.make_range('chrM', 1, 10)))
self.assertFalse(
ref.is_valid_interval(ranges.make_range('chrM', 1, 100000)))
self.assertEqual(len(ref.contigs), 3)
self.assertEqual([c.name for c in ref.contigs], chr_names)
self.assertEqual([c.n_bases for c in ref.contigs], chr_lengths)
for contig in ref.contigs:
self.assertEqual(ref.contig(contig.name), contig)
self.assertTrue(ref.has_contig(contig.name))
self.assertFalse(ref.has_contig(contig.name + '.unknown'))
def test_main(self):
examples_out = test_utils.test_tmpfile('output.tfrecord')
ngs_errors.make_ngs_error_examples(
ref_path=test_utils.genomics_core_testdata(
'ucsc.hg19.chr20.unittest.fasta.gz'),
vcf_path=test_utils.genomics_core_testdata(
'test_nist.b37_chr20_100kbp_at_10mb.vcf.gz'),
bam_path=test_utils.genomics_core_testdata(
'NA12878_S1.chr20.10_10p1mb.bam'),
examples_out_path=examples_out,
max_reads=100)
actual_examples = _read_examples(examples_out)
golden_examples = _read_examples(
test_utils.genomics_core_testdata(
'golden.examples.ngs_errors.tfrecord'))
self.assertEqual(len(actual_examples), 100)
assertExamplesAreEqual(self,
golden_examples,
actual_examples,
expected_keys={
'read_name', 'cigar', 'read_sequence',
'read_qualities', 'true_sequence'
})
def setUp(self):
self.sites_vcf = test_utils.genomics_core_testdata('test_sites.vcf')
self.samples_vcf = test_utils.genomics_core_testdata('test_samples.vcf.gz')
self.options = variants_pb2.VcfReaderOptions()
self.sites_reader = vcf_reader.VcfReader.from_file(self.sites_vcf,
self.options)
self.samples_reader = vcf_reader.VcfReader.from_file(
self.samples_vcf, self.options)
def setUp(self):
self.sites_reader = vcf.VcfReader(
test_utils.genomics_core_testdata('test_sites.vcf'),
use_index=False)
self.samples_reader = vcf.VcfReader(
test_utils.genomics_core_testdata('test_samples.vcf.gz'),
use_index=True)
def test_from_file_raises_with_missing_inputs(self, fasta_filename,
fai_filename):
fasta = test_utils.genomics_core_testdata(fasta_filename)
fai = test_utils.genomics_core_testdata(fai_filename)
with self.assertRaisesRegexp(
ValueError,
'Not found: could not load fasta and/or fai for fasta ' + fasta):
indexed_fasta_reader.IndexedFastaReader.from_file(fasta, fai)
def test_main(self):
in_ref = test_utils.genomics_core_testdata('test.fasta')
in_vcf = test_utils.genomics_core_testdata('test_phaseset.vcf')
with mock.patch.object(sys, 'exit') as mock_exit:
validate_vcf.main(['validate_vcf', in_ref, in_vcf])
# Only the first call to sys.exit() matters.
self.assertEqual(mock.call(-1), mock_exit.mock_calls[0])
def test_iterate(self, fasta_filename):
# Check the indexed fasta file's iterable matches that of the unindexed
# fasta file.
indexed_fasta_reader = fasta.IndexedFastaReader(
test_utils.genomics_core_testdata(fasta_filename))
unindexed_fasta_reader = fasta.UnindexedFastaReader(
test_utils.genomics_core_testdata(fasta_filename))
self.assertEqual(list(indexed_fasta_reader.iterate()),
list(unindexed_fasta_reader.iterate()))
def test_from_file_raises_with_missing_inputs(self, fasta_filename,
fai_filename):
fasta = test_utils.genomics_core_testdata(fasta_filename)
fai = test_utils.genomics_core_testdata(fai_filename)
with self.assertRaisesRegexp(
ValueError,
'Not found: could not load fasta and/or fai for fasta ' +
fasta):
reference_fai.GenomeReferenceFai.from_file(fasta, fai)
def setUp(self):
self.unindexed_options = variants_pb2.VcfReaderOptions()
self.indexed_options = variants_pb2.VcfReaderOptions(
index_mode=index_pb2.INDEX_BASED_ON_FILENAME)
self.sites_vcf = test_utils.genomics_core_testdata('test_sites.vcf')
self.samples_vcf = test_utils.genomics_core_testdata('test_samples.vcf.gz')
self.sites_reader = vcf_reader.VcfReader.from_file(self.sites_vcf,
self.unindexed_options)
self.samples_reader = vcf_reader.VcfReader.from_file(
self.samples_vcf, self.indexed_options)
def test_from_file_raises_with_missing_inputs(self, fasta_filename,
fai_filename):
fasta = test_utils.genomics_core_testdata(fasta_filename)
fai = test_utils.genomics_core_testdata(fai_filename)
# TODO(b/196638558): OpError exception not propagated.
with self.assertRaisesRegexp(
ValueError,
'could not load fasta and/or fai for fasta ' + fasta):
reference.IndexedFastaReader.from_file(
fasta, fai, fasta_pb2.FastaReaderOptions())
def test_dispatching_reader(self):
with fasta.FastaReader(
test_utils.genomics_core_testdata('test.fasta')) as reader:
# The reader is an instance of IndexedFastaReader which supports query().
self.assertEqual(reader.query(ranges.make_range('chrM', 1, 6)), 'ATCAC')
with fasta.FastaReader(
test_utils.genomics_core_testdata('unindexed.fasta')) as reader:
# The reader is an instance of UnindexedFastaReader which doesn't support
# query().
with self.assertRaises(NotImplementedError):
reader.query(ranges.make_range('chrM', 1, 5))
def _make_reader(self, filename, has_embedded_ref):
if has_embedded_ref:
# If we have an embedded reference, force the reader to use it by not
# providing an argument for ref_path.
return sam.SamReader(test_utils.genomics_core_testdata(filename))
else:
# Otherwise we need to explicitly override the reference encoded in the UR
# of the CRAM file to use the path provided to our test.fasta.
return sam.SamReader(
test_utils.genomics_core_testdata(filename),
ref_path=test_utils.genomics_core_testdata('test.fasta'))
def setUp(self):
tfrecord_file = test_utils.genomics_core_testdata(
'test_features.gff.tfrecord')
self.records = list(
io_utils.read_tfrecords(tfrecord_file, proto=gff_pb2.GffRecord))
self.header = gff_pb2.GffHeader(
sequence_regions=[ranges.make_range('ctg123', 0, 1497228)])
def test_roundtrip(self,
expected_infos,
expected_fmt,
expected_fmt1,
expected_fmt2,
reader_excluded_info=None,
reader_excluded_format=None,
writer_excluded_info=None,
writer_excluded_format=None):
expected_records = [
record.format(info=info, fmt=expected_fmt, efmts1=e1,
efmts2=e2) for record, info, e1, e2 in zip(
self.record_format_strings, expected_infos,
expected_fmt1, expected_fmt2)
]
expected = self.header + ''.join(expected_records)
with vcf.VcfReader(
test_utils.genomics_core_testdata('test_py_roundtrip.vcf'),
excluded_info_fields=reader_excluded_info,
excluded_format_fields=reader_excluded_format) as reader:
records = list(reader.iterate())
output_path = test_utils.test_tmpfile('test_roundtrip_tmpfile.vcf')
with vcf.VcfWriter(
output_path,
header=reader.header,
excluded_info_fields=writer_excluded_info,
excluded_format_fields=writer_excluded_format) as writer:
for record in records:
writer.write(record)
with open(output_path) as f:
actual = f.read()
self.assertEqual(actual, expected)
def test_c_reader(self):
self.assertNotEqual(self.sites_reader.c_reader, 0)
self.assertNotEqual(self.samples_reader.c_reader, 0)
tfrecord_reader = vcf.VcfReader(
test_utils.genomics_core_testdata('test_samples.vcf.golden.tfrecord'))
self.assertNotEqual(tfrecord_reader.c_reader, 0)
def setUp(self):
super(TabixTest, self).setUp()
self.input_file = test_utils.genomics_core_testdata(
'test_samples.vcf.gz')
self.output_file = test_utils.test_tmpfile('test_samples.vcf.gz')
shutil.copyfile(self.input_file, self.output_file)
self.tbx_index_file = self.output_file + '.tbi'
def test_sam_iterate_raises_on_malformed_record(self):
malformed = test_utils.genomics_core_testdata('malformed.sam')
reader = sam_reader.SamReader.from_file(malformed, self.options)
iterable = iter(reader.iterate())
self.assertIsNotNone(next(iterable))
with self.assertRaises(ValueError):
list(iterable)
def test_conversion_to_tfrecord_and_back(self, original_input_file):
"""Test conversion from a native file format to tfrecord.gz, then back."""
input_path = test_utils.genomics_core_testdata(original_input_file)
tfrecord_output_path = test_utils.test_tmpfile(original_input_file +
".tfrecord.gz")
native_output_path = test_utils.test_tmpfile(original_input_file)
# Test conversion from native format to tfrecord.
self._convert(input_path, tfrecord_output_path)
# TODO(b/63133103): remove this when SAM writer is implemented.
if native_output_path.endswith(".sam"):
raise unittest.SkipTest("SAM writing not yet supported")
# Test conversion from tfrecord format back to native format. Ensure that
# conversions where we would need a header, but don't have one from the
# input, trigger an error message.
if any(
native_output_path.endswith(ext)
for ext in FORMATS_REQUIRING_HEADER):
with self.assertRaisesRegexp(
converter.ConversionError,
"Input file does not have a header, which is needed to construct "
"output file"):
self._convert(tfrecord_output_path, native_output_path)
else:
self._convert(tfrecord_output_path, native_output_path)
def test_native_gff_header(self, gff_filename):
gff_path = test_utils.genomics_core_testdata(gff_filename)
with gff.GffReader(gff_path) as reader:
self.assertEqual(EXPECTED_GFF_VERSION, reader.header.gff_version)
with gff.NativeGffReader(gff_path) as native_reader:
self.assertEqual(EXPECTED_GFF_VERSION,
native_reader.header.gff_version)
def test_headless_sam_raises(self):
headerless = test_utils.genomics_core_testdata('headerless.sam')
with self.assertRaisesRegex(
ValueError, 'Could not parse file with bad SAM header'):
sam_reader.SamReader.from_file(reads_path=headerless,
ref_path='',
options=self.options)
def test_bed_iterate_raises_on_malformed_record(self, filename):
malformed = test_utils.genomics_core_testdata(filename)
reader = bed_reader.BedReader.from_file(malformed, self.options)
iterable = iter(reader.iterate())
self.assertIsNotNone(next(iterable))
with self.assertRaises(ValueError):
list(iterable)
def test_round_trip_vcf(self, test_datum_name):
# Round-trip variants through writing and reading:
# 1. Read variants v1 from VcfReader;
# 2. Write v1 to vcf using our VcfWriter;
# 3. Read back in using VcfReader -- v2;
# 4. compare v1 and v2.
in_file = test_utils.genomics_core_testdata(test_datum_name)
out_file = test_utils.test_tmpfile('output_' + test_datum_name)
v1_reader = vcf.VcfReader(in_file)
v1_records = list(v1_reader.iterate())
self.assertTrue(v1_records, 'Reader failed to find records')
header = copy.deepcopy(v1_reader.header)
writer_options = variants_pb2.VcfWriterOptions()
with vcf_writer.VcfWriter.to_file(out_file, header,
writer_options) as writer:
for record in v1_records:
writer.write(record)
v2_reader = vcf.VcfReader(out_file)
v2_records = list(v2_reader.iterate())
self.assertEqual(v1_records, v2_records,
'Round-tripped variants not as expected')
def test_headless_sam_raises(self):
headerless = test_utils.genomics_core_testdata('headerless.sam')
reader = sam_reader.SamReader.from_file(reads_path=headerless,
ref_path='',
options=self.options)
iterable = iter(reader.iterate())
with self.assertRaises(ValueError):
next(iterable)
def setUp(self):
self.bed = test_utils.genomics_core_testdata('test_regions.bed')
self.zipped_bed = test_utils.genomics_core_testdata(
'test_regions.bed.gz')
self.options = bed_pb2.BedReaderOptions()
self.first = bed_pb2.BedRecord(reference_name='chr1',
start=10,
end=20,
name='first',
score=100,
strand=bed_pb2.BedRecord.FORWARD_STRAND,
thick_start=12,
thick_end=18,
item_rgb='255,124,1',
block_count=3,
block_sizes='2,6,2',
block_starts='10,12,18')
def test_from_regions(self, regions, expected):
# For convenience we allow 'test.bed' in our regions but the actual file
# path is in our testdata directory.
for i in range(len(regions)):
if regions[i] == 'test.bed':
regions[i] = test_utils.genomics_core_testdata('test.bed')
self.assertEqual(list(ranges.from_regions(regions)), expected)
def test_from_bed(self, bed_filename):
source = test_utils.genomics_core_testdata(bed_filename)
self.assertCountEqual([
ranges.make_range('chr1', 1, 10),
ranges.make_range('chr2', 20, 30),
ranges.make_range('chr2', 40, 60),
ranges.make_range('chr3', 80, 90),
], ranges.RangeSet.from_bed(source))
def testCompressedExplicit(self):
reader = genomics_reader.TFRecordReader(
test_utils.genomics_core_testdata('test_features.gff.tfrecord.gz'),
gff_pb2.GffRecord(),
compression_type='GZIP')
records = list(reader.iterate())
self.assertEqual('GenBank', records[0].source)
self.assertEqual('ctg123', records[1].range.reference_name)
def test_query_on_unindexed_reader_raises(self):
window = ranges.parse_literal('chr1:10,000,000-10,000,100')
unindexed_file = test_utils.genomics_core_testdata('test_samples.vcf')
with vcf_reader.VcfReader.from_file(unindexed_file,
self.options) as reader:
with self.assertRaisesRegexp(ValueError,
'Cannot query without an index'):
reader.query(window)
def test_iterate_bed_reader(self, bed_filename):
bed_path = test_utils.genomics_core_testdata(bed_filename)
expected = [('chr1', 10, 20), ('chr1', 100, 200)]
with bed.BedReader(bed_path) as reader:
records = list(reader.iterate())
self.assertLen(records, 2)
self.assertEqual([(r.reference_name, r.start, r.end) for r in records],
expected)
