def annotate_mutation(self, mutation):
chr = mutation.chr
start = int(mutation.start)
end = int(mutation.end)
txs = self.get_transcripts_by_pos(chr, start, end)
final_annotation_dict = self._create_blank_set_of_annotations()
final_annotation_dict['variant_type'] = Annotation(value=TranscriptProviderUtils.infer_variant_type(mutation.ref_allele, mutation.alt_allele), datasourceName=self.title)
chosen_tx = None
# We have hit IGR if no transcripts come back. Most annotations can just use the blank set.
if len(txs) == 0:
final_annotation_dict['variant_classification'] = self._create_basic_annotation(VariantClassification.IGR)
nearest_genes = self._get_nearest_genes(chr, int(start), int(end))
final_annotation_dict['other_transcripts'] = self._create_basic_annotation(value='%s (%s upstream) : %s (%s downstream)' % (nearest_genes[0][0], nearest_genes[0][1], nearest_genes[1][0], nearest_genes[1][1]))
final_annotation_dict['gene'] = self._create_basic_annotation('Unknown')
final_annotation_dict['gene_id'] = self._create_basic_annotation('0')
final_annotation_dict['genome_change'] = self._create_basic_annotation(TranscriptProviderUtils.determine_genome_change(mutation.chr, mutation.start, mutation.end, mutation.ref_allele, mutation.alt_allele, final_annotation_dict['variant_type'].value))
else:
# Choose the best effect transcript
chosen_tx = self._choose_transcript(txs, self.get_tx_mode(), final_annotation_dict['variant_type'].value, mutation.ref_allele, mutation.alt_allele, start, end)
vcer = VariantClassifier()
final_annotation_dict['annotation_transcript'] = self._create_basic_annotation(chosen_tx.get_transcript_id())
final_annotation_dict['genome_change'] = self._create_basic_annotation(TranscriptProviderUtils.determine_genome_change(mutation.chr, mutation.start, mutation.end, mutation.ref_allele, mutation.alt_allele, final_annotation_dict['variant_type'].value))
final_annotation_dict['strand'] = self._create_basic_annotation(chosen_tx.get_strand())
final_annotation_dict['transcript_position'] = self._create_basic_annotation(TranscriptProviderUtils.render_transcript_position(int(start), int(end), chosen_tx))
final_annotation_dict['transcript_id'] = self._create_basic_annotation(chosen_tx.get_transcript_id())
variant_classfication = vcer.variant_classify(tx=chosen_tx, variant_type=final_annotation_dict['variant_type'].value,
ref_allele=mutation.ref_allele, alt_allele=mutation.alt_allele, start=mutation.start, end=mutation.end)
final_annotation_dict['transcript_exon'] = self._create_basic_annotation(str(variant_classfication.get_exon_i()+1))
final_annotation_dict['variant_classification'] = self._create_basic_annotation(variant_classfication.get_vc())
final_annotation_dict['secondary_variant_classification'] = self._create_basic_annotation(variant_classfication.get_secondary_vc())
final_annotation_dict['protein_change'] = self._create_basic_annotation(vcer.generate_protein_change_from_vc(variant_classfication))
final_annotation_dict['codon_change'] = self._create_basic_annotation(vcer.generate_codon_change_from_vc(chosen_tx, start, end, variant_classfication))
final_annotation_dict['transcript_change'] = self._create_basic_annotation(vcer.generate_transcript_change_from_tx(chosen_tx, final_annotation_dict['variant_type'].value, variant_classfication, start, end, mutation.ref_allele, mutation.alt_allele))
final_annotation_dict['transcript_strand'] = self._create_basic_annotation(chosen_tx.get_strand())
final_annotation_dict['gene'] = self._create_basic_annotation(chosen_tx.get_gene())
final_annotation_dict['gene_type'] = self._create_basic_annotation(chosen_tx.get_gene_type())
final_annotation_dict['gencode_transcript_tags'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'tag'))
final_annotation_dict['gencode_transcript_status'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'transcript_status'))
final_annotation_dict['havana_transcript'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'havana_transcript'))
final_annotation_dict['ccds_id'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'ccdsid'))
final_annotation_dict['gencode_transcript_type'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'transcript_type'))
final_annotation_dict['gencode_transcript_name'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'transcript_name'))
other_transcript_value = self._render_other_transcripts(txs, [txs.index(chosen_tx)], final_annotation_dict['variant_type'].value, mutation.ref_allele, mutation.alt_allele, mutation.start, mutation.end)
final_annotation_dict['other_transcripts'] = self._create_basic_annotation(other_transcript_value)
# final_annotation_dict['gene_id'].value
mutation.addAnnotations(final_annotation_dict)
# Add the HGVS annotations ... setting to "" if not available.
hgvs_dict_annotations = self._create_hgvs_annotation_dict(mutation, chosen_tx)
mutation.addAnnotations(hgvs_dict_annotations)
return mutation
def _render_other_transcripts(self, txs, transcriptIndicesToSkip, variant_type, ref_allele, alt_allele, start, end):
"""
Create a list of transcripts that are not being chosen.
Other transcripts are formatted <gene>_<transcript_id>_<variant_classification>_<protein_change>
Note: There are other areas of Oncotator (e.g. Generic_GeneProteinPositionDatasource) that depend
on this format. Changing it here may introduce bugs in other pieces of code.
Also, do not include any transcript that would render as IGR.
txs -- a list of transcripts to render.
transcriptIndicesToSkip -- a list of transcripts that are being used (i.e. not an "other transcript"). This will usually be the canonical or any transcript chosen by tx_mode.
"""
vcer = VariantClassifier()
other_transcripts = list()
for i, ot in enumerate(txs):
if i not in transcriptIndicesToSkip:
vc = vcer.variant_classify(tx=ot, variant_type=variant_type, ref_allele=ref_allele, alt_allele=alt_allele, start=start, end=end)
if vc.get_vc() == VariantClassification.IGR:
continue
o = '_'.join([ot.get_gene(), ot.get_transcript_id(),
vc.get_vc(), vcer.generate_protein_change_from_vc(vc)])
o = o.strip('_')
other_transcripts.append(o)
return '|'.join(other_transcripts)
def _render_other_transcripts(self, txs, transcriptIndicesToSkip, variant_type, ref_allele, alt_allele, start, end):
"""
Create a list of transcripts that are not being chosen.
Other transcripts are formatted <gene>_<transcript_id>_<variant_classification>_<protein_change>
Note: There are other areas of Oncotator (e.g. Generic_GeneProteinPositionDatasource) that depend
on this format. Changing it here may introduce bugs in other pieces of code.
Also, do not include any transcript that would render as IGR.
txs -- a list of transcripts to render.
transcriptIndicesToSkip -- a list of transcripts that are being used (i.e. not an "other transcript"). This will usually be the canonical or any transcript chosen by tx_mode.
"""
vcer = VariantClassifier()
other_transcripts = list()
for i, ot in enumerate(txs):
if i not in transcriptIndicesToSkip:
vc = vcer.variant_classify(tx=ot, variant_type=variant_type, ref_allele=ref_allele, alt_allele=alt_allele, start=start, end=end)
if vc.get_vc() == VariantClassification.IGR:
continue
o = '_'.join([ot.get_gene(), ot.get_transcript_id(),
vc.get_vc(), vcer.generate_protein_change_from_vc(vc)])
o = o.strip('_')
other_transcripts.append(o)
return '|'.join(other_transcripts)
def _choose_best_effect_transcript(self, txs, variant_type, ref_allele, alt_allele, start, end):
"""Choose the transcript with the most detrimental effect.
The rankings are in TranscriptProviderUtils.
Ties are broken by which transcript has the longer coding length.
:param list txs: list of Transcript
:param str variant_type:
:param str ref_allele:
:param str alt_allele:
:param str start:
:param str end:
:return Transcript:
"""
vcer = VariantClassifier()
effect_dict = TranscriptProviderUtils.retrieve_effect_dict()
best_effect_score = 100000000 # lower score is more likely to get picked
best_effect_tx = None
for tx in txs:
if (ref_allele == "" or ref_allele == "-") and (alt_allele == "" or alt_allele == "-"):
vc = VariantClassification.SILENT
else:
vc = vcer.variant_classify(tx, ref_allele, alt_allele, start, end, variant_type).get_vc()
effect_score = effect_dict.get(vc, 25)
if effect_score < best_effect_score:
best_effect_score = effect_score
best_effect_tx = tx
elif (effect_score == best_effect_score) and (len(best_effect_tx.get_seq()) < len(tx.get_seq())):
best_effect_score = effect_score
best_effect_tx = tx
return best_effect_tx
def annotate_mutation(self, mutation):
chr = mutation.chr
start = int(mutation.start)
end = int(mutation.end)
txs = self.get_transcripts_by_pos(chr, start, end)
final_annotation_dict = self._create_blank_set_of_annotations()
final_annotation_dict['variant_type'] = Annotation(value=TranscriptProviderUtils.infer_variant_type(mutation.ref_allele, mutation.alt_allele), datasourceName=self.title)
chosen_tx = None
# We have hit IGR if no transcripts come back. Most annotations can just use the blank set.
if len(txs) == 0:
final_annotation_dict['variant_classification'] = self._create_basic_annotation(VariantClassification.IGR)
nearest_genes = self._get_nearest_genes(chr, int(start), int(end))
final_annotation_dict['other_transcripts'] = self._create_basic_annotation(value='%s (%s upstream) : %s (%s downstream)' % (nearest_genes[0][0], nearest_genes[0][1], nearest_genes[1][0], nearest_genes[1][1]))
final_annotation_dict['gene'] = self._create_basic_annotation('Unknown')
final_annotation_dict['gene_id'] = self._create_basic_annotation('0')
final_annotation_dict['genome_change'] = self._create_basic_annotation(TranscriptProviderUtils.determine_genome_change(mutation.chr, mutation.start, mutation.end, mutation.ref_allele, mutation.alt_allele, final_annotation_dict['variant_type'].value))
else:
# Choose the best effect transcript
chosen_tx = self._choose_transcript(txs, self.get_tx_mode(), final_annotation_dict['variant_type'].value, mutation.ref_allele, mutation.alt_allele, start, end)
vcer = VariantClassifier()
final_annotation_dict['annotation_transcript'] = self._create_basic_annotation(chosen_tx.get_transcript_id())
final_annotation_dict['genome_change'] = self._create_basic_annotation(TranscriptProviderUtils.determine_genome_change(mutation.chr, mutation.start, mutation.end, mutation.ref_allele, mutation.alt_allele, final_annotation_dict['variant_type'].value))
final_annotation_dict['strand'] = self._create_basic_annotation(chosen_tx.get_strand())
final_annotation_dict['transcript_position'] = self._create_basic_annotation(TranscriptProviderUtils.render_transcript_position(int(start), int(end), chosen_tx))
final_annotation_dict['transcript_id'] = self._create_basic_annotation(chosen_tx.get_transcript_id())
variant_classfication = vcer.variant_classify(tx=chosen_tx, variant_type=final_annotation_dict['variant_type'].value,
ref_allele=mutation.ref_allele, alt_allele=mutation.alt_allele, start=mutation.start, end=mutation.end)
final_annotation_dict['transcript_exon'] = self._create_basic_annotation(str(variant_classfication.get_exon_i()+1))
final_annotation_dict['variant_classification'] = self._create_basic_annotation(variant_classfication.get_vc())
final_annotation_dict['secondary_variant_classification'] = self._create_basic_annotation(variant_classfication.get_secondary_vc())
final_annotation_dict['protein_change'] = self._create_basic_annotation(vcer.generate_protein_change_from_vc(variant_classfication))
final_annotation_dict['codon_change'] = self._create_basic_annotation(vcer.generate_codon_change_from_vc(chosen_tx, start, end, variant_classfication))
final_annotation_dict['transcript_change'] = self._create_basic_annotation(vcer.generate_transcript_change_from_tx(chosen_tx, final_annotation_dict['variant_type'].value, variant_classfication, start, end, mutation.ref_allele, mutation.alt_allele))
final_annotation_dict['transcript_strand'] = self._create_basic_annotation(chosen_tx.get_strand())
final_annotation_dict['gene'] = self._create_basic_annotation(chosen_tx.get_gene())
final_annotation_dict['gene_type'] = self._create_basic_annotation(chosen_tx.get_gene_type())
final_annotation_dict['gencode_transcript_tags'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'tag'))
final_annotation_dict['gencode_transcript_status'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'transcript_status'))
final_annotation_dict['havana_transcript'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'havana_transcript'))
final_annotation_dict['ccds_id'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'ccdsid'))
final_annotation_dict['gencode_transcript_type'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'transcript_type'))
final_annotation_dict['gencode_transcript_name'] = self._create_basic_annotation(self._retrieve_gencode_tag_value(chosen_tx, 'transcript_name'))
other_transcript_value = self._render_other_transcripts(txs, [txs.index(chosen_tx)], final_annotation_dict['variant_type'].value, mutation.ref_allele, mutation.alt_allele, mutation.start, mutation.end)
final_annotation_dict['other_transcripts'] = self._create_basic_annotation(other_transcript_value)
# final_annotation_dict['gene_id'].value
mutation.addAnnotations(final_annotation_dict)
# Add the HGVS annotations ... setting to "" if not available.
hgvs_dict_annotations = self._create_hgvs_annotation_dict(mutation, chosen_tx)
mutation.addAnnotations(hgvs_dict_annotations)
return mutation
def _calculate_effect_score(tx, start, end, alt_allele, ref_allele, variant_type):
"""Compute the effect score"""
effect_dict = TranscriptProviderUtils.retrieve_effect_dict()
vcer = VariantClassifier()
if (ref_allele == "" or ref_allele == "-") and (alt_allele == "" or alt_allele == "-"):
vc = VariantClassification.SILENT
else:
vc = vcer.variant_classify(tx, ref_allele, alt_allele, start, end, variant_type).get_vc()
effect_score = effect_dict.get(vc, 25)
return effect_score
def _calculate_effect_score(tx, start, end, alt_allele, ref_allele, variant_type):
"""Compute the effect score"""
effect_dict = TranscriptProviderUtils.retrieve_effect_dict()
vcer = VariantClassifier()
if (ref_allele == "" or ref_allele == "-") and (alt_allele == "" or alt_allele == "-"):
vc = VariantClassification.SILENT
else:
vc = vcer.variant_classify(tx, ref_allele, alt_allele, start, end, variant_type).get_vc()
effect_score = effect_dict.get(vc, 25)
return effect_score