def test_should_add_short_sequence_and_quality_list(self):
sample_bank = SampleBank("AAA")
sample_bank.add_sample_with_seqs_and_quals("SAMPLE1", ["...", "007"])
read_lists = [
builder.build_reads(0, {}) for builder in sample_bank["SAMPLE1"]
]
reads = [read for read_list in read_lists for read in read_list]
self.assertEqual(reads[0].qual, "!!g")
def test_should_add_seq_and_quals_list_with_deletion(self):
sample_bank = SampleBank("AAA")
sample_bank.add_sample_with_seqs_and_quals("SAMPLE1", [".*C", "1 3"])
read_lists = [
builder.build_reads(0, {}) for builder in sample_bank["SAMPLE1"]
]
reads = [read for read_list in read_lists for read in read_list]
self.assertEqual(reads[0].qual, "+?")
def test_should_add_seq_and_quals_list_with_fwd_and_rev_reads(self):
sample_bank = SampleBank("AAA")
sample_bank.add_sample_with_seqs_and_quals("SAMPLE1", ["...", "007"],
n_fwd=1,
n_rev=2)
self.assertEqual(len(sample_bank["SAMPLE1"]), 1)
self.assertEqual(sample_bank["SAMPLE1"][0].n_fwd, 1)
self.assertEqual(sample_bank["SAMPLE1"][0].n_rev, 2)
def __run_wecall_variant_caller(self, chrom, reference_string, sequence_list, vcf_stem=None):
if vcf_stem is None:
vcf_stem = chrom
sample_bank = SampleBank(reference_string, chrom=chrom)
sample_bank.add_sample_with_seqs_and_quals(DEFAULT_SAMPLE_NAME, sequence_list, n_fwd=10, n_rev=10)
vc_builder = VariantCallerBuilderFromSampleBank(sample_bank, self.work_dir)
vc_wrapper = vc_builder.build()
vc_wrapper.add_additional_command("allowMNPCalls", False)
vc_wrapper.output_vcf = path.join(self.intermediate_vcfs_dir, "{}.vcf".format(vcf_stem))
vc_wrapper.run()
return vc_wrapper.output_vcf
def __build_default_sample_bank(
self,
ref,
sequence_list,
n_fwd=None,
n_rev=None):
sample_bank = SampleBank(ref)
sample_bank.add_sample_with_seqs_and_quals(
DEFAULT_SAMPLE_NAME, sequence_list, n_fwd, n_rev)
return sample_bank
def test_should_add_two_sequence_list(self):
sample_bank = SampleBank("AAA")
sample_bank.add_sample_with_seqs_and_quals("SAMPLE1", ["...", " .."])
read_lists = [
builder.build_reads(0, {}) for builder in sample_bank["SAMPLE1"]
]
reads = [read for read_list in read_lists for read in read_list]
reads.sort(key=lambda x: (x.pos, x.seq, x.qual, x.cigarstring, x.mapq))
self.assertEqual(len(reads), 2)
self.assertEqual(reads[0].qual, "HHH")
self.assertEqual(reads[1].qual, "HH")
def test_should_add_complex_seq_and_quals_list(self):
sample_bank = SampleBank("AAA")
sample_bank.add_sample_with_seqs_and_quals(
"SAMPLE1", ["...", "007", " ..", ".*C", "1 3"])
read_lists = [
builder.build_reads(0, {}) for builder in sample_bank["SAMPLE1"]
]
reads = [read for read_list in read_lists for read in read_list]
reads.sort(key=lambda x: (x.pos, x.seq, x.qual, x.cigarstring, x.mapq))
self.assertEqual(len(reads), 3)
self.assertEqual(reads[0].qual, "!!g")
self.assertEqual(reads[1].qual, "+?")
self.assertEqual(reads[2].qual, self.default_char * 2)
def assert_quality_recalibrated_in_output_bam(self, ref_string,
input_bam_seqs,
output_bam_seqs):
input_sample_bank = SampleBank(ref_string)
input_sample_bank.add_sample_with_seqs_and_quals(
self.sample_name, input_bam_seqs)
output_sample_bank = SampleBank(ref_string)
output_sample_bank.add_sample_with_seqs_and_quals(
self.sample_name, output_bam_seqs)
vc_builder = VariantCallerBuilderFromSampleBank(
input_sample_bank, self.work_dir)
vc_builder.configuration["recalibrateBaseQs"] = "true"
vc_builder.configuration[
"intermediateRecalibFileStem"] = self.output_stem
vc_builder.build().run()
self.assertTrue(os.path.exists(self.output_sam))
sam_file = pysam.Samfile(self.output_sam, "r")
reads = list(sam_file.fetch())
self.assertEqual(len(reads), len(output_sample_bank[self.sample_name]))
# Sort the sam as in sequence bank.
# output_sample_bank.sort_sequence_banks()
output_reads = sorted(output_sample_bank[self.sample_name].build_reads(
0, {}),
key=lambda x:
(x.pos, x.seq, x.qual, x.cigarstring, x.mapq))
reads.sort(key=lambda x: (x.pos, x.seq, x.qual, x.cigarstring, x.mapq))
for read, expected_sequence in zip(reads, output_reads):
self.assertEqual(read.pos, expected_sequence.pos)
self.assertEqual(read.seq, expected_sequence.seq)
self.assert_matching_ascii_qualities(read.qual,
expected_sequence.qual)
self.assertEqual(read.cigarstring, expected_sequence.cigarstring)
self.assertEqual(read.mapq, expected_sequence.mapq)
sam_file.close()
def calls_variants(self, ref, sequence_list, candidate_ascii_haplotypes, prior, expected_ascii_haplotypes):
sample_bank = SampleBank(ref)
sample_bank.add_sample_with_seqs_and_quals("TEST", sequence_list, 1, 0)
variant_generator = AsciiVariantGenerator(sample_bank.reference)
candidate_variants = variant_generator.get_variants(candidate_ascii_haplotypes)
expected_variants = variant_generator.get_variants(expected_ascii_haplotypes)
candidate_variant_list = VCFBuilder(path.join(self.work_dir, "candiate_variants.vcf"))
candidate_variant_list.schema.set_info_data('AF', 'A', 'Float', 'Allele Frequency')
for var in candidate_variants:
candidate_variant_list.with_record_from_variant(
var, info=InfoData(candidate_variant_list.schema, {"AF": prior})
)
candidate_variant_list.build().index()
vc_wrapper_builder = VariantCallerBuilderFromSampleBank(sample_bank, self.work_dir)
vc_wrapper_builder.configuration[CANDIDATE_VARIANTS_FILE_KEY] = candidate_variant_list.compressed_filename
callset = vc_wrapper_builder.build().run().get_variant_callset(self)
self.assertEqual(callset.get_variants(), set(expected_variants))
def calls_variants_from_samples(self,
ref,
sample_seqs,
expected_haplotypes=None,
expected_call_stubs=None,
config_dict=None):
"""
:param expected_haplotypes: dictionary: {sample_name : list of two ascii sequences expressing the genotype}
:param expected_call_stubs: dictionary: {variant_stub: dictionary {sample_name: str(genotype)} }
"""
self.__validate_expected_calls(
expected_haplotypes, expected_call_stubs)
sample_bank = SampleBank(ref)
for sample_name, sequence_list in sample_seqs.items():
sample_bank.add_sample_with_seqs_and_quals(
sample_name, sequence_list)
variant_callset = self.__run_wecall(sample_bank, config_dict)
wecall_calls = variant_callset.get_variants_with_genotypes()
if expected_call_stubs is None:
self.__filter_none_genotypes(wecall_calls)
expected_calls = self.__get_expected_calls_from_sample_ascii_haplotypes(
expected_haplotypes, sample_bank.reference)
else:
expected_calls = {}
for variant_stub, genotypes in expected_call_stubs.items():
variant = self._variant_from_stub(
sample_bank.reference.chrom, variant_stub)
expected_calls[variant] = OrderedDict()
for sample_name, genotype in genotypes.items():
expected_calls[variant][sample_name] = GenotypeCall(
genotype)
self.maxDiff = None # print the whole message if the following assertion fails
self.assertDictEqual(expected_calls, wecall_calls)
