def variants_to_protein_sequences_dataframe(
expressed_vcf="data/b16.f10/b16.expressed.vcf",
not_expressed_vcf="data/b16.f10/b16.not-expressed.vcf",
tumor_rna_bam="data/b16.f10/b16.combined.sorted.bam",
min_mapping_quality=0,
max_protein_sequences_per_variant=1,
variant_sequence_assembly=False):
"""
Helper function to load pair of VCFs and tumor RNA BAM
and use them to generate a DataFrame of expressed variant protein
sequences.
"""
expressed_variants = load_vcf(expressed_vcf)
not_expressed_variants = load_vcf(not_expressed_vcf)
combined_variants = VariantCollection(
list(expressed_variants) + list(not_expressed_variants))
samfile = load_bam(tumor_rna_bam)
allele_reads_generator = reads_overlapping_variants(
variants=combined_variants,
samfile=samfile,
min_mapping_quality=min_mapping_quality)
protein_sequences_generator = reads_generator_to_protein_sequences_generator(
allele_reads_generator,
max_protein_sequences_per_variant=max_protein_sequences_per_variant,
variant_sequence_assembly=variant_sequence_assembly)
df = protein_sequences_generator_to_dataframe(protein_sequences_generator)
return df, expressed_variants, combined_variants
def isovar_protein_sequence_dict(self):
"""
This function computes a dictionary of Variant objects to a single isovar protein sequence
that will be used to try to construct VaccinePeptides. If this function has been previously
called, the result will be cached.
"""
if self._isovar_protein_sequence_dict is None:
# total number of amino acids is the vaccine peptide length plus the
# number of off-center windows around the mutation
protein_fragment_sequence_length = (
self.vaccine_peptide_length + 2 * self.padding_around_mutation)
"""
These sequences are only the ones that overlap the variant and support the mutation.
Right now, this generator yields:
- (variant, mutant protein sequences) if there's enough alt RNA support
- (variant, None) if the variant is silent or there are ref reads overlapping the
variant locus but inadequate alt RNA support.
- does not return the variant if there's no RNA support for ref or alt - we may miss
some coding variants this way unless we check for them explicitly
Future intended behavior: returns all passing variants, with a protein sequences
generator that is non empty if there are enough alt RNA reads supporting the variant
"""
protein_sequences_generator = reads_generator_to_protein_sequences_generator(
self.reads_generator,
transcript_id_whitelist=None,
protein_sequence_length=protein_fragment_sequence_length,
min_alt_rna_reads=self.min_alt_rna_reads,
min_variant_sequence_coverage=self.
min_variant_sequence_coverage,
variant_sequence_assembly=self.variant_sequence_assembly,
max_protein_sequences_per_variant=1)
self._isovar_protein_sequence_dict = {}
for variant, isovar_protein_sequences in protein_sequences_generator:
if len(isovar_protein_sequences) == 0:
# variant RNA support is below threshold
logger.info("No protein sequences for %s", variant)
continue
# use the first protein sequence - why?
self._isovar_protein_sequence_dict[
variant] = isovar_protein_sequences[0]
return self._isovar_protein_sequence_dict
def test_variants_to_protein_sequences_dataframe_filtered_all_reads_by_mapping_quality():
# since the B16 BAM has all MAPQ=255 values then all the reads should get dropped
# if we set the minimum quality to 256
variants = load_vcf("data/b16.f10/b16.vcf")
samfile = load_bam("data/b16.f10/b16.combined.sorted.bam")
allele_reads_generator = reads_overlapping_variants(
variants=variants,
samfile=samfile,
min_mapping_quality=256)
protein_sequences_generator = reads_generator_to_protein_sequences_generator(
allele_reads_generator,
max_protein_sequences_per_variant=1)
df = protein_sequences_generator_to_dataframe(protein_sequences_generator)
print(df)
eq_(
len(df),
0,
"Expected 0 entries, got %d: %s" % (len(df), df))
def isovar_protein_sequence_dict(self):
"""
This function computes a dictionary of Variant objects to a single isovar protein sequence
that will be used to try to construct VaccinePeptides. If this function has been previously
called, the result will be cached.
"""
if self._isovar_protein_sequence_dict is None:
# total number of amino acids is the vaccine peptide length plus the
# number of off-center windows around the mutation
protein_fragment_sequence_length = (
self.vaccine_peptide_length + 2 * self.padding_around_mutation)
"""
These sequences are only the ones that overlap the variant and support the mutation.
Right now, this generator yields:
- (variant, mutant protein sequences) if there's enough alt RNA support
- (variant, None) if the variant is silent or there are ref reads overlapping the
variant locus but inadequate alt RNA support.
- does not return the variant if there's no RNA support for ref or alt - we may miss
some coding variants this way unless we check for them explicitly
Future intended behavior: returns all passing variants, with a protein sequences
generator that is non empty if there are enough alt RNA reads supporting the variant
"""
protein_sequences_generator = reads_generator_to_protein_sequences_generator(
self.reads_generator,
transcript_id_whitelist=None,
protein_sequence_length=protein_fragment_sequence_length,
min_alt_rna_reads=self.min_alt_rna_reads,
min_variant_sequence_coverage=self.min_variant_sequence_coverage,
variant_sequence_assembly=self.variant_sequence_assembly,
max_protein_sequences_per_variant=1)
self._isovar_protein_sequence_dict = {}
for variant, isovar_protein_sequences in protein_sequences_generator:
if len(isovar_protein_sequences) == 0:
# variant RNA support is below threshold
logger.info("No protein sequences for %s", variant)
continue
# use the first protein sequence - why?
self._isovar_protein_sequence_dict[variant] = isovar_protein_sequences[0]
return self._isovar_protein_sequence_dict
def generate_vaccine_peptides(
reads_generator,
mhc_predictor,
vaccine_peptide_length,
padding_around_mutation,
max_vaccine_peptides_per_variant,
min_alt_rna_reads,
min_variant_sequence_coverage,
variant_sequence_assembly,
min_epitope_score=0):
"""
Returns dictionary mapping each variant to list of VaccinePeptide objects.
"""
# total number of amino acids is the vaccine peptide length plus the
# number of off-center windows around the mutation
protein_fragment_sequence_length = (
vaccine_peptide_length + 2 * padding_around_mutation)
protein_sequences_generator = reads_generator_to_protein_sequences_generator(
reads_generator,
transcript_id_whitelist=None,
protein_sequence_length=protein_fragment_sequence_length,
min_alt_rna_reads=min_alt_rna_reads,
min_variant_sequence_coverage=min_variant_sequence_coverage,
variant_sequence_assembly=variant_sequence_assembly,
max_protein_sequences_per_variant=1)
result_dict = {}
for variant, isovar_protein_sequences in protein_sequences_generator:
vaccine_peptides = vaccine_peptides_for_variant(
variant=variant,
isovar_protein_sequences=isovar_protein_sequences,
mhc_predictor=mhc_predictor,
vaccine_peptide_length=vaccine_peptide_length,
padding_around_mutation=padding_around_mutation,
max_vaccine_peptides_per_variant=max_vaccine_peptides_per_variant,
min_epitope_score=min_epitope_score)
result_dict[variant] = vaccine_peptides
return result_dict
def generate_vaccine_peptides(
reads_generator,
mhc_predictor,
vaccine_peptide_length,
padding_around_mutation,
max_vaccine_peptides_per_variant,
min_alt_rna_reads,
min_variant_sequence_coverage,
variant_sequence_assembly,
num_mutant_epitopes_to_keep=10000,
min_epitope_score=0):
"""
Returns a tuple of two values:
- dictionary mapping each variant to list of VaccinePeptide objects
- dictionary containing some variant counts for report display
"""
# total number of amino acids is the vaccine peptide length plus the
# number of off-center windows around the mutation
protein_fragment_sequence_length = (
vaccine_peptide_length + 2 * padding_around_mutation)
protein_sequences_generator = reads_generator_to_protein_sequences_generator(
reads_generator,
transcript_id_whitelist=None,
protein_sequence_length=protein_fragment_sequence_length,
min_alt_rna_reads=min_alt_rna_reads,
min_variant_sequence_coverage=min_variant_sequence_coverage,
variant_sequence_assembly=variant_sequence_assembly,
max_protein_sequences_per_variant=1)
result_dict = {}
counts_dict = defaultdict(int)
for variant, isovar_protein_sequences in protein_sequences_generator:
if len(variant.effects().drop_silent_and_noncoding()) > 0:
counts_dict['num_coding_effect_variants'] += 1
isovar_protein_sequences = list(isovar_protein_sequences)
if len(isovar_protein_sequences) == 0:
# this means the variant RNA support is below threshold
logger.info("No protein sequences for %s", variant)
continue
# use the first protein sequence - why?
counts_dict['num_variants_with_rna_support'] += 1
isovar_protein_sequence = isovar_protein_sequences[0]
vaccine_peptides = vaccine_peptides_for_variant(
variant=variant,
isovar_protein_sequence=isovar_protein_sequence,
mhc_predictor=mhc_predictor,
vaccine_peptide_length=vaccine_peptide_length,
padding_around_mutation=padding_around_mutation,
max_vaccine_peptides_per_variant=max_vaccine_peptides_per_variant,
num_mutant_epitopes_to_keep=num_mutant_epitopes_to_keep,
min_epitope_score=min_epitope_score)
# do any of this variant's vaccine peptides contain mutant epitopes?
any_mutant_epitopes = False
for vaccine_peptide in vaccine_peptides:
if vaccine_peptide.contains_mutant_epitopes():
any_mutant_epitopes = True
break
if any_mutant_epitopes:
counts_dict['num_variants_with_vaccine_peptides'] += 1
result_dict[variant] = vaccine_peptides
for key, value in counts_dict.items():
logger.info('%s: %d', key, value)
return result_dict, counts_dict
