def test_variants_to_protein_sequences_dataframe_protein_sequence_length():
expressed_variants = load_vcf("data/b16.f10/b16.expressed.vcf")
parser = make_protein_sequences_arg_parser()
parser.print_help()
for desired_length in range(9, 20, 3):
args = parser.parse_args([
"--vcf", data_path("data/b16.f10/b16.vcf"),
"--bam", data_path("data/b16.f10/b16.combined.sorted.bam"),
"--max-protein-sequences-per-variant", "1",
"--protein-sequence-length", str(desired_length),
])
df = protein_sequences_dataframe_from_args(args)
eq_(
len(df),
len(expressed_variants),
"Expected %d entries for protein_sequence_length=%d, got %d results: %s" % (
len(expressed_variants),
desired_length,
len(df),
df))
protein_sequences = df["amino_acids"]
print(protein_sequences)
protein_sequence_lengths = protein_sequences.str.len()
assert (protein_sequence_lengths == desired_length).all(), (
protein_sequence_lengths,)
def test_isovar_result_str():
for result in run_isovar(
variants=data_path("data/b16.f10/b16.vcf"),
alignment_file=data_path("data/b16.f10/b16.combined.sorted.bam")):
s = str(result)
assert len(s) > 0
assert s.startswith("IsovarResult(")
assert s.endswith(")")
def test_isovar_main_to_dataframe():
results = run_isovar(
variants=data_path("data/b16.f10/b16.vcf"),
alignment_file=data_path("data/b16.f10/b16.combined.sorted.bam"))
df = isovar_results_to_dataframe(results)
print(df)
eq_(len(df), 4)
# B16 test data has 2/4 variants with enough coverage
# to translate protein sequences
eq_(df["passes_all_filters"].sum(), 2)
def test_isovar_result_property_types():
for result in run_isovar(
variants=data_path("data/b16.f10/b16.vcf"),
alignment_file=data_path("data/b16.f10/b16.combined.sorted.bam")):
# variant
assert type(result.variant) is Variant
# counts of genes and transcripts from variant
assert type(result.num_overlapping_genes) is int
assert type(result.num_overlapping_coding_genes) is int
assert type(result.num_overlapping_transcripts) is int
assert type(result.num_overlapping_coding_transcripts) is int
# protein sequence
assert type(result.top_protein_sequence) in (type(None),
ProteinSequence)
# counts of genes and transcripts from protein sequences
assert type(result.num_genes_from_protein_sequences) is int
assert type(result.num_genes_from_top_protein_sequence) is int
assert type(result.num_transcripts_from_protein_sequences) is int
assert type(result.num_transcripts_from_top_protein_sequence) is int
# read and fragment counts
assert type(result.num_ref_reads) is int
assert type(result.num_alt_reads) is int
assert type(result.num_other_reads) is int
assert type(result.num_ref_fragments) is int
assert type(result.num_alt_fragments) is int
assert type(result.num_other_fragments) is int
# read and fragment fractions
assert type(result.fraction_ref_reads) is float
assert type(result.fraction_alt_reads) is float
assert type(result.fraction_other_reads) is float
assert type(result.fraction_ref_fragments) is float
assert type(result.fraction_alt_fragments) is float
assert type(result.fraction_other_fragments) is float
# read and fragment count ratios
assert type(result.ratio_alt_to_other_reads) is float
assert type(result.ratio_alt_to_other_fragments) is float
assert type(result.ratio_other_to_alt_fragments) is float
assert type(result.ratio_other_to_alt_reads) is float
assert type(result.ratio_ref_to_other_fragments) is float
assert type(result.ratio_other_to_ref_fragments) is float
assert type(result.ratio_other_to_ref_reads) is float
# this property aggregates all filters
assert result.passes_all_filters in {True, False}
assert type(result.protein_sequence_mutation_start) in (int,
type(None))
assert type(result.protein_sequence_mutation_end) in (int, type(None))
def test_locus_reads_dataframe():
sam_all_variants = load_bam("data/b16.f10/b16.combined.bam")
n_reads_expected = 0
sam_path_single_variant = data_path(
"data/b16.f10/b16.f10.127a.aldh1b1.chr4.45802539.refG.altC.sam")
with open(sam_path_single_variant) as f:
for line in f:
if line.startswith("HWI"):
n_reads_expected += 1
# we know from inspecting the file that *one* of the reads overlapping this
# variant has a CIGAR string of N at the location before and thus we'll
# be missing that read.
#
# TODO: figure out what to do when the variant nucleotide is at the start or
# end of an exon, since that won't have mapping positions on both its left
# and right
n_reads_expected -= 1
print("Found %d sequences in %s" % (n_reads_expected, sam_path_single_variant))
df = locus_reads_dataframe(
samfile=sam_all_variants,
chromosome="chr4",
base1_position_before_variant=45802538,
base1_position_after_variant=45802540)
print(df)
eq_(len(df), n_reads_expected)
def test_locus_reads_dataframe():
sam_all_variants = load_bam("data/b16.f10/b16.combined.bam")
n_reads_expected = 0
sam_path_single_variant = data_path(
"data/b16.f10/b16.f10.127a.aldh1b1.chr4.45802539.refG.altC.sam")
with open(sam_path_single_variant) as f:
for line in f:
if line.startswith("HWI"):
n_reads_expected += 1
# we know from inspecting the file that *one* of the reads overlapping this
# variant has a CIGAR string of N at the location before and thus we'll
# be missing that read.
#
# TODO: figure out what to do when the variant nucleotide is at the start or
# end of an exon, since that won't have mapping positions on both its left
# and right
n_reads_expected -= 1
print("Found %d sequences in %s" %
(n_reads_expected, sam_path_single_variant))
df = locus_reads_dataframe(samfile=sam_all_variants,
chromosome="chr4",
base1_position_before_variant=45802538,
base1_position_after_variant=45802540)
print(df)
eq_(len(df), n_reads_expected)
def test_mutant_amino_acids_in_mm10_chrX_8125624_refC_altA_pS460I():
# there are two co-occurring variants in the RNAseq data but since
# they don't happen in the same codon then we're considering the Varcode
# annotation to be correct
# TODO: deal with phasing of variants explicitly so that both
# variant positions are considered mutated
parser = make_protein_sequences_arg_parser()
args = parser.parse_args([
"--vcf", data_path("data/b16.f10/b16.f10.Wdr13.vcf"),
"--bam", data_path("data/b16.f10/b16.combined.sorted.bam"),
"--max-protein-sequences-per-variant", "1",
"--protein-sequence-length", "15"
])
for variant, protein_sequences in protein_sequences_generator_from_args(args):
protein_sequence = protein_sequences[0]
check_mutant_amino_acids(variant, protein_sequence)
def test_mutant_amino_acids_in_mm10_chrX_8125624_refC_altA_pS460I():
# there are two co-occurring variants in the RNAseq data but since
# they don't happen in the same codon then we're considering the Varcode
# annotation to be correct
# TODO: deal with phasing of variants explicitly so that both
# variant positions are considered mutated
parser = make_protein_sequences_arg_parser()
args = parser.parse_args([
"--vcf", data_path("data/b16.f10/b16.f10.Wdr13.vcf"),
"--bam", data_path("data/b16.f10/b16.combined.sorted.bam"),
"--max-protein-sequences-per-variant", "1",
"--protein-sequence-length", "15"
])
for variant, protein_sequences in protein_sequences_generator_from_args(args):
protein_sequence = protein_sequences[0]
check_mutant_amino_acids(variant, protein_sequence)
def test_mutant_amino_acids_in_mm10_chr9_82927102_refG_altT_pT441H():
# the variant chr9:82927102 G>T occurs right next to T>G so the varcode
# prediction for the protein sequence (Asparagine) will be wrong since
# the correct translation is Histidine
parser = make_protein_sequences_arg_parser()
args = parser.parse_args([
"--vcf", data_path("data/b16.f10/b16.f10.Phip.vcf"),
"--bam", data_path("data/b16.f10/b16.combined.sorted.bam"),
"--max-protein-sequences-per-variant", "1",
"--protein-sequence-length", "15"
])
for variant, protein_sequences in protein_sequences_generator_from_args(args):
protein_sequence = protein_sequences[0]
check_mutant_amino_acids(
variant,
protein_sequence,
expected_amino_acids="H")
def test_mutant_amino_acids_in_mm10_chr9_82927102_refG_altT_pT441H():
# the variant chr9:82927102 G>T occurs right next to T>G so the varcode
# prediction for the protein sequence (Asparagine) will be wrong since
# the correct translation is Histidine
parser = make_protein_sequences_arg_parser()
args = parser.parse_args([
"--vcf", data_path("data/b16.f10/b16.f10.Phip.vcf"),
"--bam", data_path("data/b16.f10/b16.combined.sorted.bam"),
"--max-protein-sequences-per-variant", "1",
"--protein-sequence-length", "15"
])
for variant, protein_sequences in protein_sequences_generator_from_args(args):
protein_sequence = protein_sequences[0]
check_mutant_amino_acids(
variant,
protein_sequence,
expected_amino_acids="H")
def test_isovar_result_nonsyn_variants():
for result in run_isovar(
variants=data_path("data/b16.f10/b16.vcf"),
alignment_file=data_path("data/b16.f10/b16.combined.sorted.bam")):
print(result.variant)
print(result.predicted_effect)
if result.has_mutant_protein_sequence_from_rna:
assert result.num_amino_acid_mismatches_from_predicted_effect is not None
assert result.num_amino_acid_mismatches_from_reference is not None
assert result.num_amino_acid_mismatches_from_reference > 0
eq_(result.protein_sequence_matches_reference, False)
eq_(result.protein_sequence_contains_mutation, True)
else:
assert result.num_amino_acid_mismatches_from_predicted_effect is None
assert result.num_amino_acid_mismatches_from_reference is None
assert result.protein_sequence_matches_predicted_mutation_effect is None
assert result.protein_sequence_matches_reference is None
assert result.protein_sequence_contains_mutation is None
import tempfile
from os import remove
from os.path import getsize, exists
from testing_helpers import data_path
from isovar.cli.isovar_translations import run as isovar_translations
from isovar.cli.isovar_allele_counts import run as isovar_allele_counts
from isovar.cli.isovar_allele_reads import run as isovar_allele_reads
from isovar.cli.isovar_protein_sequences import run as isovar_protein_sequences
from isovar.cli.isovar_reference_contexts import run as isovar_reference_contexts
from isovar.cli.isovar_variant_reads import run as isovar_variant_reads
from isovar.cli.isovar_variant_sequences import run as isovar_variant_sequences
from isovar.cli.isovar_main import run as isovar_main
vcf_args = ["--vcf", data_path("data/b16.f10/b16.vcf")]
args_with_bam = vcf_args + [
"--bam", data_path("data/b16.f10/b16.combined.sorted.bam")
]
def run_cli_fn(fn, include_bam_in_args=True):
with tempfile.NamedTemporaryFile(delete=False) as f:
output_path = f.name
assert not exists(output_path) == 0
output_args = ["--output", output_path]
if include_bam_in_args:
args = args_with_bam + output_args
else:
args = vcf_args + output_args
def test_isovar_result_clone_with_updates():
for result in run_isovar(
variants=data_path("data/b16.f10/b16.vcf"),
alignment_file=data_path("data/b16.f10/b16.combined.sorted.bam")):
result2 = result.clone_with_updates(variant=None)
assert result != result2
def test_isovar_result_clone():
for result in run_isovar(
variants=data_path("data/b16.f10/b16.vcf"),
alignment_file=data_path("data/b16.f10/b16.combined.sorted.bam")):
result2 = result.clone()
eq_(result, result2)
