def generate_transcript_change_from_tx(self, tx, variant_type, vc, start_genomic_space, end_genomic_space, ref_allele, alt_allele):
"""
:param vc:
:return:
"""
if vc.get_vc() == VariantClassification.SPLICE_SITE and vc.get_secondary_vc() == VariantClassification.INTRON:
return ""
# dist_from_exon = self._get_splice_site_coordinates(tx, start_genomic_space, end_genomic_space, vc.get_exon_i())
# exon_i = vc.get_exon_i()
# return TranscriptProviderUtils.render_splice_site_transcript_change(tx, dist_from_exon, exon_i, vc.get_secondary_vc() == VariantClassification.INTRON)
if vc.get_cds_start_in_exon_space() == "" or vc.get_cds_start_in_exon_space() < 0:
return ""
exon_position_start,exon_position_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(int(start_genomic_space), int(end_genomic_space), tx)
if tx.get_strand() == "-":
cds_position_start_cds_space = exon_position_start - int(vc.get_cds_start_in_exon_space())+1
cds_position_end_cds_space = exon_position_end - int(vc.get_cds_start_in_exon_space())+1
else:
cds_position_start_cds_space = exon_position_start - int(vc.get_cds_start_in_exon_space())
cds_position_end_cds_space = exon_position_end - int(vc.get_cds_start_in_exon_space())
observed_allele_stranded, reference_allele_stranded = self._get_stranded_alleles(ref_allele, alt_allele, tx)
result = TranscriptProviderUtils.render_transcript_change(variant_type, vc.get_vc(), cds_position_start_cds_space, cds_position_end_cds_space, reference_allele_stranded, observed_allele_stranded, vc.get_secondary_vc())
return result
def test_seq(self, start, end, gt):
"""Test that we can successfully determine the codon at an arbitrary location on test transcript"""
tx = self.retrieve_test_transcript_MAPK1()
transcript_position_start, transcript_position_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(start, end, tx)
transcript_seq = tx.get_seq()
seq = transcript_seq[transcript_position_start:transcript_position_end+1]
self.assertTrue(seq == gt, "Incorrect seq found guess,gt (%s, %s)" %(seq, gt))
def test_codon_single_base(self, start, end, ref_base_stranded, gt_codon):
"""Test that we can grab the proper three bases of a codon for an arbitrary single base """
tx = self.retrieve_test_transcript_MAPK1()
transcript_position_start, transcript_position_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(start, end, tx)
cds_start, cds_stop = TranscriptProviderUtils.determine_cds_in_exon_space(tx)
protein_position_start, protein_position_end = TranscriptProviderUtils.get_protein_positions(transcript_position_start, transcript_position_end, cds_start)
cds_codon_start, cds_codon_end = TranscriptProviderUtils.get_cds_codon_positions(protein_position_start, protein_position_end, cds_start)
codon_seq = tx.get_seq()[cds_codon_start:cds_codon_end+1]
self.assertTrue(codon_seq == gt_codon, "Did not get correct codon (%s): %s loc: %s-%s" %(gt_codon, codon_seq, start, end))
def test_convert_genomic_space_to_exon_space(self, loc, gt_d):
"""Test genomic --> exon transform on real data. """
gencode_input_gtf = "testdata/gencode/MAPK1.gencode.v18.annotation.gtf"
gencode_input_fasta = "testdata/gencode/MAPK1.gencode.v18.pc_transcripts.fa"
base_output_filename = "out/test_variant_classification"
shutil.rmtree(base_output_filename + ".transcript.idx", ignore_errors=True)
shutil.rmtree(base_output_filename + ".transcript_by_gene.idx", ignore_errors=True)
shutil.rmtree(base_output_filename + ".transcript_by_gp_bin.idx", ignore_errors=True)
genome_build_factory = GenomeBuildFactory()
genome_build_factory.construct_ensembl_indices([gencode_input_gtf], [gencode_input_fasta], base_output_filename)
ensembl_ds = EnsemblTranscriptDatasource(base_output_filename, version="TEST")
tx = ensembl_ds.get_overlapping_transcripts("22", "22108790", "22108790")
start, end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(loc[0], loc[1], tx[0])
loc_length = (int(loc[1]) - int(loc[0]))
self.assertTrue((end - start) == loc_length, str(end) + " - " + str(start) + " was not correct length: " + str(loc_length))
self.assertTrue(start == gt_d, "start position (" + str(start) + ") did not match gt (" + str(end) + ")" + " exons: " + str(tx[0].get_exons()))
def test_querying_transcripts_by_region(self):
"""Test web api backend call /transcripts/.... """
datasource_list = DatasourceFactory.createDatasources(
self._determine_db_dir(), "hg19", isMulticore=False)
annotator = Annotator()
for ds in datasource_list:
annotator.addDatasource(ds)
txs = annotator.retrieve_transcripts_by_region("4", 50164411, 60164411)
self.assertTranscriptsFound(txs)
## Here is an example of getting enough data to populate the json in doc/transcript_json_commented.json.txt
# None of these values are validated.
for tx in txs:
transcript_id = tx.get_transcript_id()
tx_start = tx.determine_transcript_start()
tx_end = tx.determine_transcript_stop()
gene = tx.get_gene()
chr = tx.get_contig()
n_exons = len(tx.get_exons())
strand = tx.get_strand()
footprint_start, footprint_end = tx.determine_cds_footprint()
klass = tx.get_gene_type()
cds_start = tx.determine_cds_start()
cds_end = tx.determine_cds_stop()
id = tx.get_gene_id()
genomic_coords = [[exon[0], exon[1]] for exon in tx.get_exons()]
transcript_coords = [[
TranscriptProviderUtils.convert_genomic_space_to_exon_space(
exon[0] + 1, exon[1], tx)
] for exon in tx.get_exons()]
code_len = int(cds_end) - int(cds_start) + 1
# If refseq datasources are not available, this will fail.
# Step 2 annotate the transcript, which produces a dummy mutation with the refseq annotations.
dummy_mut = annotator.annotate_transcript(tx)
refseq_mRNA_id = dummy_mut["gencode_xref_refseq_mRNA_id"]
refseq_prot_id = dummy_mut["gencode_xref_refseq_prot_acc"]
# Description is unavailable right now
description = ""
self.assertTrue(refseq_mRNA_id is not None)
self.assertTrue(refseq_prot_id is not None)
self.assertTrue(len(transcript_coords) == n_exons)
def test_querying_transcripts_by_region(self):
"""Test web api backend call /transcripts/.... """
datasource_list = DatasourceFactory.createDatasources(self._determine_db_dir(), "hg19", isMulticore=False)
annotator = Annotator()
for ds in datasource_list:
annotator.addDatasource(ds)
txs = annotator.retrieve_transcripts_by_region("4", 50164411, 60164411)
self.assertTranscriptsFound(txs)
## Here is an example of getting enough data to populate the json in doc/transcript_json_commented.json.txt
# None of these values are validated.
for tx in txs:
transcript_id = tx.get_transcript_id()
tx_start = tx.determine_transcript_start()
tx_end = tx.determine_transcript_stop()
gene = tx.get_gene()
chr = tx.get_contig()
n_exons = len(tx.get_exons())
strand = tx.get_strand()
footprint_start, footprint_end = tx.determine_cds_footprint()
klass = tx.get_gene_type()
cds_start = tx.determine_cds_start()
cds_end = tx.determine_cds_stop()
id = tx.get_gene_id()
genomic_coords = [[exon[0], exon[1]] for exon in tx.get_exons()]
transcript_coords = [
[TranscriptProviderUtils.convert_genomic_space_to_exon_space(exon[0] + 1, exon[1], tx)]
for exon in tx.get_exons()
]
code_len = int(cds_end) - int(cds_start) + 1
# If refseq datasources are not available, this will fail.
# Step 2 annotate the transcript, which produces a dummy mutation with the refseq annotations.
dummy_mut = annotator.annotate_transcript(tx)
refseq_mRNA_id = dummy_mut["gencode_xref_refseq_mRNA_id"]
refseq_prot_id = dummy_mut["gencode_xref_refseq_prot_acc"]
# Description is unavailable right now
description = ""
self.assertTrue(refseq_mRNA_id is not None)
self.assertTrue(refseq_prot_id is not None)
self.assertTrue(len(transcript_coords) == n_exons)
def generate_transcript_change_from_tx(self, tx, variant_type, vc,
start_genomic_space,
end_genomic_space, ref_allele,
alt_allele):
"""
:param vc:
:return:
"""
if vc.get_vc(
) == VariantClassification.SPLICE_SITE and vc.get_secondary_vc(
) == VariantClassification.INTRON:
return ""
# dist_from_exon = self._get_splice_site_coordinates(tx, start_genomic_space, end_genomic_space, vc.get_exon_i())
# exon_i = vc.get_exon_i()
# return TranscriptProviderUtils.render_splice_site_transcript_change(tx, dist_from_exon, exon_i, vc.get_secondary_vc() == VariantClassification.INTRON)
if vc.get_cds_start_in_exon_space(
) == "" or vc.get_cds_start_in_exon_space() < 0:
return ""
exon_position_start, exon_position_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(
int(start_genomic_space), int(end_genomic_space), tx)
if tx.get_strand() == "-":
cds_position_start_cds_space = exon_position_start - int(
vc.get_cds_start_in_exon_space()) + 1
cds_position_end_cds_space = exon_position_end - int(
vc.get_cds_start_in_exon_space()) + 1
else:
cds_position_start_cds_space = exon_position_start - int(
vc.get_cds_start_in_exon_space())
cds_position_end_cds_space = exon_position_end - int(
vc.get_cds_start_in_exon_space())
observed_allele_stranded, reference_allele_stranded = self._get_stranded_alleles(
ref_allele, alt_allele, tx)
result = TranscriptProviderUtils.render_transcript_change(
variant_type, vc.get_vc(), cds_position_start_cds_space,
cds_position_end_cds_space, reference_allele_stranded,
observed_allele_stranded, vc.get_secondary_vc())
return result
def variant_classify(self, tx, ref_allele, alt_allele, start, end, variant_type, dist=2):
"""Perform classifications.
Everything handled in genomic space
*RNA*
x'UTR
Splice_Site (Intron)
Intron
Splice_Site (Exon)
{Missense, Silent}
{Nonsense, Silent}
{Nonstop, Silent}
IGR
x'Flank
De_novo_Start
"""
gene_type = tx.get_gene_type()
if gene_type != "protein_coding":
if gene_type == VariantClassification.LINCRNA:
return VariantClassification(VariantClassification.LINCRNA, variant_type, tx.get_transcript_id())
else:
return VariantClassification(VariantClassification.RNA, variant_type, tx.get_transcript_id())
if ref_allele == "-":
ref_allele = ""
if alt_allele == "-":
alt_allele = ""
s = int(start)
e = int(end)
is_exon_overlap = TranscriptProviderUtils.determine_if_exon_overlap(s, e, tx, variant_type)
is_splice_site_tuple = self._determine_if_splice_site_overlap(s, e, tx, variant_type, dist)
is_splice_site = is_splice_site_tuple[0]
is_beyond_exons, side, is_flank = self._determine_beyond_exon_info_vt(start, end, tx, variant_type)
if not is_exon_overlap and not is_beyond_exons:
exon_i = TranscriptProviderUtils.determine_closest_exon(tx, int(start), int(end))
if is_splice_site:
# Intron Splice Site
return VariantClassification(VariantClassification.SPLICE_SITE, variant_type, tx.get_transcript_id(), vc_secondary=VariantClassification.INTRON, exon_i=exon_i)
else:
return VariantClassification(VariantClassification.INTRON, variant_type, tx.get_transcript_id(), exon_i=exon_i)
if not is_exon_overlap and is_beyond_exons:
if is_flank:
# Flanks
if side.startswith("3"):
return VariantClassification(VariantClassification.THREE_PRIME_PRIME_FLANK, variant_type, transcript_id=tx.get_transcript_id())
else:
return VariantClassification(VariantClassification.FIVE_PRIME_PRIME_FLANK, variant_type, transcript_id=tx.get_transcript_id())
else:
# IGR
return VariantClassification(VariantClassification.IGR, variant_type)
is_start_codon_overlap = self._determine_codon_overlap(s, e, tx.get_start_codon(), variant_type)
is_stop_codon_overlap = self._determine_codon_overlap(s, e, tx.get_stop_codon(), variant_type)
if is_start_codon_overlap and not variant_type.endswith("NP"):
return VariantClassification('Start_Codon_' + variant_type.capitalize(), variant_type, transcript_id=tx.get_transcript_id())
if is_stop_codon_overlap and not variant_type.endswith("NP"):
return VariantClassification('Stop_Codon_' + variant_type.capitalize(), variant_type, transcript_id=tx.get_transcript_id())
is_cds_overlap = self._determine_if_cds_overlap(s, e, tx, variant_type)
if is_exon_overlap and not is_cds_overlap and not is_start_codon_overlap and not is_stop_codon_overlap:
# UTR
if side.startswith("3"):
vc_tmp = VariantClassification.THREE_PRIME_UTR
else:
vc_tmp = VariantClassification.FIVE_PRIME_UTR
transcript_position_exon_space_start, transcript_position_exon_space_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(start, end, tx)
vc = self._determine_de_novo(vc_tmp, transcript_position_exon_space_start, ref_allele, alt_allele, tx, variant_type)
return VariantClassification(vc, variant_type, transcript_id=tx.get_transcript_id(), )
# We have a clean overlap in the CDS. Includes start codon or stop codon.
if is_cds_overlap or is_stop_codon_overlap or is_start_codon_overlap:
is_frameshift_indel = self.is_frameshift_indel(variant_type, int(start), int(end), alt_allele)
return self._determine_vc_for_cds_overlap(start, end, ref_allele, alt_allele, is_frameshift_indel, is_splice_site, tx, variant_type, is_start_codon_overlap)
raise ValueError("Could not determine variant classification: " + tx.get_trancript_id() + " " + str([ref_allele, alt_allele, start, end]))
def _determine_vc_for_cds_overlap(self, start, end, ref_allele, alt_allele, is_frameshift_indel, is_splice_site, tx, variant_type, is_start_codon):
"""
Note: This method can also handle start and stop codons.
:param start:
:param end:
:param ref_allele:
:param alt_allele:
:param is_frameshift_indel:
:param is_splice_site:
:param tx:
:param variant_type:
:return:
"""
observed_allele_stranded, reference_allele_stranded = self._get_stranded_alleles(ref_allele, alt_allele, tx)
transcript_position_start, transcript_position_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(
start, end, tx)
if tx.get_strand() == "+" and not variant_type == VariantClassification.VT_INS:
transcript_position_start -= 1
transcript_position_end -= 1
transcript_seq = tx.get_seq()
protein_seq = tx.get_protein_seq()
cds_start, cds_stop = TranscriptProviderUtils.determine_cds_in_exon_space(tx)
protein_position_start, protein_position_end = TranscriptProviderUtils.get_protein_positions(
transcript_position_start,
transcript_position_end, cds_start)
new_ref_transcript_seq = transcript_seq
if (transcript_seq[transcript_position_start:transcript_position_end+1] != reference_allele_stranded) and variant_type != VariantClassification.VT_INS:
new_ref_transcript_seq = list(transcript_seq)
new_ref_transcript_seq[transcript_position_start:transcript_position_end+1] = reference_allele_stranded
new_ref_transcript_seq = ''.join(new_ref_transcript_seq)
ref_tx_seq_has_been_changed = True
else:
ref_tx_seq_has_been_changed = False
cds_codon_start, cds_codon_end = TranscriptProviderUtils.get_cds_codon_positions(protein_position_start, protein_position_end, cds_start)
if variant_type == "DEL":
reference_codon_seq = new_ref_transcript_seq[cds_codon_start:cds_codon_end+1].lower()
else:
reference_codon_seq = TranscriptProviderUtils.mutate_reference_sequence(new_ref_transcript_seq[cds_codon_start:cds_codon_end+1].lower(), cds_codon_start, transcript_position_start, transcript_position_end, reference_allele_stranded, variant_type)
if variant_type == "INS" and tx.get_strand() == "-":
mutated_codon_seq = TranscriptProviderUtils.mutate_reference_sequence(reference_codon_seq.lower(), cds_codon_start - 1, transcript_position_start, transcript_position_end, observed_allele_stranded, variant_type)
else:
mutated_codon_seq = TranscriptProviderUtils.mutate_reference_sequence(reference_codon_seq.lower(), cds_codon_start, transcript_position_start, transcript_position_end, observed_allele_stranded, variant_type)
observed_aa = Bio.Seq.translate(mutated_codon_seq)
if ref_tx_seq_has_been_changed:
reference_aa = Bio.Seq.translate(reference_codon_seq)
else:
reference_aa = protein_seq[protein_position_start-1:protein_position_end]
if variant_type != VariantClassification.VT_SNP:
try:
reference_aa, observed_aa, protein_position_start, protein_position_end = \
self._adjust_protein_position_and_alleles(protein_seq, protein_position_start,
protein_position_end, reference_aa, observed_aa)
except InvalidVariantException as ive:
logging.getLogger(__name__).error("Could not properly adjust protein position for variant: %s, %s, %s, %s, %s VT: %s" % (tx.get_contig(), start, end, ref_allele, alt_allele, variant_type))
logging.getLogger(__name__).error(str(ive))
logging.getLogger(__name__).warn("Above error may not have exact start and end positions if this is a VCF input.")
logging.getLogger(__name__).warn("Variant type is likely incorrect. This can happen with some GATK VCFs")
logging.getLogger(__name__).warn(TranscriptProviderUtils.is_valid_xNP(variant_type, ref_allele, alt_allele))
logging.getLogger(__name__).warn("The protein_change annotation may not be properly rendered.")
vc_tmp, vc_tmp_secondary = self.infer_variant_classification(variant_type, reference_aa, observed_aa, ref_allele, alt_allele,
is_frameshift_indel=is_frameshift_indel, is_splice_site=is_splice_site, is_start_codon=is_start_codon)
cds_start_exon_space, cds_end_exon_space = TranscriptProviderUtils.determine_cds_in_exon_space(tx)
exon_i = TranscriptProviderUtils.determine_exon_index(int(start), int(end), tx, variant_type)
final_vc = VariantClassification(vc_tmp, variant_type, transcript_id=tx.get_transcript_id(), alt_codon=mutated_codon_seq, ref_codon=reference_codon_seq, ref_aa=reference_aa, ref_protein_start=protein_position_start, ref_protein_end=protein_position_end, alt_aa=observed_aa, alt_codon_start_in_exon=cds_codon_start, alt_codon_end_in_exon=cds_codon_end, ref_codon_start_in_exon=cds_codon_start, ref_codon_end_in_exon=cds_codon_end, cds_start_in_exon_space=cds_start_exon_space, ref_allele_stranded=reference_allele_stranded, alt_allele_stranded=observed_allele_stranded, exon_i=exon_i, vc_secondary=vc_tmp_secondary)
return final_vc
def variant_classify(self,
tx,
ref_allele,
alt_allele,
start,
end,
variant_type,
dist=2):
"""Perform classifications.
Everything handled in genomic space
*RNA*
x'UTR
Splice_Site (Intron)
Intron
Splice_Site (Exon)
{Missense, Silent}
{Nonsense, Silent}
{Nonstop, Silent}
IGR
x'Flank
De_novo_Start
"""
gene_type = tx.get_gene_type()
if gene_type != "protein_coding":
if gene_type == VariantClassification.LINCRNA:
return VariantClassification(VariantClassification.LINCRNA,
variant_type,
tx.get_transcript_id())
else:
return VariantClassification(VariantClassification.RNA,
variant_type,
tx.get_transcript_id())
if ref_allele == "-":
ref_allele = ""
if alt_allele == "-":
alt_allele = ""
s = int(start)
e = int(end)
is_exon_overlap = TranscriptProviderUtils.determine_if_exon_overlap(
s, e, tx, variant_type)
is_splice_site_tuple = self._determine_if_splice_site_overlap(
s, e, tx, variant_type, dist)
is_splice_site = is_splice_site_tuple[0]
is_beyond_exons, side, is_flank = self._determine_beyond_exon_info_vt(
start, end, tx, variant_type)
if not is_exon_overlap and not is_beyond_exons:
exon_i = TranscriptProviderUtils.determine_closest_exon(
tx, int(start), int(end))
if is_splice_site:
# Intron Splice Site
return VariantClassification(
VariantClassification.SPLICE_SITE,
variant_type,
tx.get_transcript_id(),
vc_secondary=VariantClassification.INTRON,
exon_i=exon_i)
else:
return VariantClassification(VariantClassification.INTRON,
variant_type,
tx.get_transcript_id(),
exon_i=exon_i)
if not is_exon_overlap and is_beyond_exons:
if is_flank:
# Flanks
if side.startswith("3"):
return VariantClassification(
VariantClassification.THREE_PRIME_PRIME_FLANK,
variant_type,
transcript_id=tx.get_transcript_id())
else:
return VariantClassification(
VariantClassification.FIVE_PRIME_PRIME_FLANK,
variant_type,
transcript_id=tx.get_transcript_id())
else:
# IGR
return VariantClassification(VariantClassification.IGR,
variant_type)
is_start_codon_overlap = self._determine_codon_overlap(
s, e, tx.get_start_codon(), variant_type)
is_stop_codon_overlap = self._determine_codon_overlap(
s, e, tx.get_stop_codon(), variant_type)
if is_start_codon_overlap and not variant_type.endswith("NP"):
return VariantClassification('Start_Codon_' +
variant_type.capitalize(),
variant_type,
transcript_id=tx.get_transcript_id())
if is_stop_codon_overlap and not variant_type.endswith("NP"):
return VariantClassification('Stop_Codon_' +
variant_type.capitalize(),
variant_type,
transcript_id=tx.get_transcript_id())
is_cds_overlap = self._determine_if_cds_overlap(s, e, tx, variant_type)
if is_exon_overlap and not is_cds_overlap and not is_start_codon_overlap and not is_stop_codon_overlap:
# UTR
if side.startswith("3"):
vc_tmp = VariantClassification.THREE_PRIME_UTR
else:
vc_tmp = VariantClassification.FIVE_PRIME_UTR
transcript_position_exon_space_start, transcript_position_exon_space_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(
start, end, tx)
vc = self._determine_de_novo(vc_tmp,
transcript_position_exon_space_start,
ref_allele, alt_allele, tx,
variant_type)
return VariantClassification(
vc,
variant_type,
transcript_id=tx.get_transcript_id(),
)
# We have a clean overlap in the CDS. Includes start codon or stop codon.
if is_cds_overlap or is_stop_codon_overlap or is_start_codon_overlap:
is_frameshift_indel = self.is_frameshift_indel(
variant_type, int(start), int(end), alt_allele)
return self._determine_vc_for_cds_overlap(
start, end, ref_allele, alt_allele, is_frameshift_indel,
is_splice_site, tx, variant_type, is_start_codon_overlap)
raise ValueError("Could not determine variant classification: " +
tx.get_trancript_id() + " " +
str([ref_allele, alt_allele, start, end]))
def _determine_vc_for_cds_overlap(self, start, end, ref_allele, alt_allele,
is_frameshift_indel, is_splice_site, tx,
variant_type, is_start_codon):
"""
Note: This method can also handle start and stop codons.
:param start:
:param end:
:param ref_allele:
:param alt_allele:
:param is_frameshift_indel:
:param is_splice_site:
:param tx:
:param variant_type:
:return:
"""
observed_allele_stranded, reference_allele_stranded = self._get_stranded_alleles(
ref_allele, alt_allele, tx)
transcript_position_start, transcript_position_end = TranscriptProviderUtils.convert_genomic_space_to_exon_space(
start, end, tx)
if tx.get_strand(
) == "+" and not variant_type == VariantClassification.VT_INS:
transcript_position_start -= 1
transcript_position_end -= 1
transcript_seq = tx.get_seq()
protein_seq = tx.get_protein_seq()
cds_start, cds_stop = TranscriptProviderUtils.determine_cds_in_exon_space(
tx)
protein_position_start, protein_position_end = TranscriptProviderUtils.get_protein_positions(
transcript_position_start, transcript_position_end, cds_start)
new_ref_transcript_seq = transcript_seq
if (transcript_seq[transcript_position_start:transcript_position_end +
1] != reference_allele_stranded
) and variant_type != VariantClassification.VT_INS:
new_ref_transcript_seq = list(transcript_seq)
new_ref_transcript_seq[
transcript_position_start:transcript_position_end +
1] = reference_allele_stranded
new_ref_transcript_seq = ''.join(new_ref_transcript_seq)
ref_tx_seq_has_been_changed = True
else:
ref_tx_seq_has_been_changed = False
cds_codon_start, cds_codon_end = TranscriptProviderUtils.get_cds_codon_positions(
protein_position_start, protein_position_end, cds_start)
if variant_type == "DEL":
reference_codon_seq = new_ref_transcript_seq[
cds_codon_start:cds_codon_end + 1].lower()
else:
reference_codon_seq = TranscriptProviderUtils.mutate_reference_sequence(
new_ref_transcript_seq[cds_codon_start:cds_codon_end +
1].lower(), cds_codon_start,
transcript_position_start, transcript_position_end,
reference_allele_stranded, variant_type)
if variant_type == "INS" and tx.get_strand() == "-":
mutated_codon_seq = TranscriptProviderUtils.mutate_reference_sequence(
reference_codon_seq.lower(), cds_codon_start - 1,
transcript_position_start, transcript_position_end,
observed_allele_stranded, variant_type)
else:
mutated_codon_seq = TranscriptProviderUtils.mutate_reference_sequence(
reference_codon_seq.lower(), cds_codon_start,
transcript_position_start, transcript_position_end,
observed_allele_stranded, variant_type)
observed_aa = MutUtils.translate_sequence(mutated_codon_seq)
if ref_tx_seq_has_been_changed:
reference_aa = MutUtils.translate_sequence(reference_codon_seq)
else:
reference_aa = protein_seq[protein_position_start -
1:protein_position_end]
if variant_type != VariantClassification.VT_SNP:
try:
reference_aa, observed_aa, protein_position_start, protein_position_end = \
self._adjust_protein_position_and_alleles(protein_seq, protein_position_start,
protein_position_end, reference_aa, observed_aa)
except InvalidVariantException as ive:
logging.getLogger(__name__).error(
"Could not properly adjust protein position for variant: %s, %s, %s, %s, %s VT: %s"
% (tx.get_contig(), start, end, ref_allele, alt_allele,
variant_type))
logging.getLogger(__name__).error(str(ive))
logging.getLogger(__name__).warn(
"Above error may not have exact start and end positions if this is a VCF input."
)
logging.getLogger(__name__).warn(
"Variant type is likely incorrect. This can happen with some GATK VCFs"
)
logging.getLogger(__name__).warn(
TranscriptProviderUtils.is_valid_xNP(
variant_type, ref_allele, alt_allele))
logging.getLogger(__name__).warn(
"The protein_change annotation may not be properly rendered."
)
vc_tmp, vc_tmp_secondary = self.infer_variant_classification(
variant_type,
reference_aa,
observed_aa,
ref_allele,
alt_allele,
is_frameshift_indel=is_frameshift_indel,
is_splice_site=is_splice_site,
is_start_codon=is_start_codon)
cds_start_exon_space, cds_end_exon_space = TranscriptProviderUtils.determine_cds_in_exon_space(
tx)
exon_i = TranscriptProviderUtils.determine_exon_index(
int(start), int(end), tx, variant_type)
final_vc = VariantClassification(
vc_tmp,
variant_type,
transcript_id=tx.get_transcript_id(),
alt_codon=mutated_codon_seq,
ref_codon=reference_codon_seq,
ref_aa=reference_aa,
ref_protein_start=protein_position_start,
ref_protein_end=protein_position_end,
alt_aa=observed_aa,
alt_codon_start_in_exon=cds_codon_start,
alt_codon_end_in_exon=cds_codon_end,
ref_codon_start_in_exon=cds_codon_start,
ref_codon_end_in_exon=cds_codon_end,
cds_start_in_exon_space=cds_start_exon_space,
ref_allele_stranded=reference_allele_stranded,
alt_allele_stranded=observed_allele_stranded,
exon_i=exon_i,
vc_secondary=vc_tmp_secondary)
return final_vc
