def pubtator_results_for_lex(lex, gene_name=None):
""" Takes an LVG object ("lex") (metavariant.VariantLVG) and uses each
variant found in lex.variants to do a search in PubTator for associated PMIDs.
Returns a dictionary of results mapping hgvs_text to PMIDs found -- i.e.:
{ hgvs_text: {'components': VariantComponents object,
'pmids': [<pmids>]
}
}
:param lex: lexical variant object (metavariant.VariantLVG)
:return: dictionary of results
"""
gene_id = None
if gene_name:
gene_id = GeneID(gene_name)
else:
try:
gene_name = lex.gene_name
gene_id = GeneID(lex.gene_name)
except TypeError:
# no gene_name? it happens -- but our results will be basically bunk without it.
return []
log.info('[%s] %s (Gene ID: %s)', lex.seqvar, gene_name, gene_id)
results = {}
for seqtype in lex.variants:
for seqvar in lex.variants[seqtype].values():
try:
result = pubtator_results_for_seqvar(seqvar, gene_id)
results.update(result)
try:
for row in results['%s' % seqvar]:
log.info(
'[%s] [[%s]] Mentions: %s PMID: %s Components: %s',
lex.seqvar, seqvar, row['Mentions'], row['PMID'],
row['Components'])
except Exception as error:
print(error)
#from IPython import embed; embed()
except RejectedSeqVar:
log.debug(
'[%s] [[%s]] VariantComponents raised RejectedSeqVar',
lex.seqvar, seqvar)
except PubtatorDBError as error:
log.info('[%s] [[%s]] %r', lex.seqvar, seqvar, error)
return results
def pubtator_lex_to_pmid(lex, gene_name=None):
""" Takes an LVG object ("lex") (metavariant.VariantLVG) and uses each
variant found in lex.variants to do a search in PubTator for associated PMIDs.
Returns a dictionary of results mapping VariantComponents objects to PMIDs found -- i.e.:
{ hgvs_text: {'comp': VariantComponents object,
'pmids': [<pmids>]
}
:param lex: lexical variant object (metavariant.VariantLVG)
:return: dictionary of results
"""
gene_id = None
if gene_name:
gene_id = GeneID(gene_name)
else:
try:
gene_name = lex.gene_name
gene_id = GeneID(lex.gene_name)
except TypeError:
# no gene_name? it happens -- but our results will be basically bunk without it.
return []
log.info('[%s] %s (Gene ID: %s)', lex.seqvar, gene_name, gene_id)
pmids = set()
for seqtype in lex.variants:
for seqvar in lex.variants[seqtype].values():
try:
components = VariantComponents(seqvar)
except RejectedSeqVar:
log.debug('[%s] Rejected sequence variant: %r' %
(lex.seqvar, seqvar))
continue
log.info('[%s] [[%s]] %s', lex.seqvar, seqvar, components)
try:
if seqtype == 'p':
results = pubtator_db.search_proteins(components, gene_id)
else:
results = pubtator_db.search_m2p(components, gene_id)
for res in results:
pmids.add(res['PMID'])
except PubtatorDBError as error:
log.info('[%s] (%s) %r', lex.seqvar, seqvar, error)
return list(pmids)
def __init__(self, gene_id=None, gene_name=None):
if gene_id:
self.gene_id = gene_id
self.gene_name = GeneName(gene_id)
elif gene_name:
self.gene_name = gene_name
self.gene_id = GeneID(gene_name)
def search_aminoDBs(gene, achg):
print('[%s]' % achg)
comp = VariantComponents(aminochange=achg)
if not comp:
print('[%s] INVALID Amino Change' % achg)
return
print('[%s] Posedit: %s' % (achg, comp.posedit))
print('[%s] Slang: %r' % (achg, comp.posedit_slang))
gene_id = GeneID(gene)
print('[%s] Gene: %s (ID: %i)' % (achg, gene, gene_id))
#results = cvdb.search(comp, gene_id, strict=False)
#print('[%s] Clinvar LOOSE matches: %r' % (achg, results))
results = cvdb.search(comp, gene_id, strict=True)
print('[%s] Clinvar STRICT matches: %i' % (achg, len(results)))
for res in results:
print('[%s]' % achg, res['PMID'], res['HGVS'], res['VariationID'], res['GeneSymbol'], res['Ref'], res['Pos'], res['Alt'])
results = pubdb.search_proteins(comp, gene_id)
print('[%s] PubtatorDB matches: %i' % (achg, len(results)))
for res in results:
print(res)
def process_hgvs_through_pubtator(hgvs_text):
print()
print('[%s]' % hgvs_text)
lex = LVG(hgvs_text)
edittype = VariantComponents(lex.seqvar).edittype
if edittype not in ['SUB', 'DEL', 'INS', 'FS', 'INDEL']:
print('[%s] Cannot process edit type %s; skipping' %
(hgvs_text, edittype))
return None
try:
gene_id = GeneID(lex.gene_name)
except TypeError:
# no gene_name? it happens.
gene_id = None
print('[%s]' % hgvs_text, lex.gene_name, '(Gene ID: %s)' % gene_id)
pmids = set()
for seqtype in lex.variants:
for seqvar in lex.variants[seqtype]:
try:
components = VariantComponents(seqvar)
except RejectedSeqVar:
print('[%s] Rejected sequence variant: %r' %
(hgvs_text, seqvar))
continue
print('[%s]' % hgvs_text, seqtype, components)
if seqtype == 'p':
results = pubtator_db.search_proteins(components, gene_id)
else:
results = pubtator_db.search_m2p(components, gene_id)
for res in results:
pmids.add(res['PMID'])
return pmids
def hgvs_to_pmid_results_dict(hgvs_text):
print()
print('[%s]' % hgvs_text)
lex = LVG(hgvs_text)
edittype = VariantComponents(lex.seqvar).edittype
if edittype not in ['SUB', 'DEL', 'INS', 'FS', 'INDEL']:
print('[%s] Cannot process edit type %s; skipping' %
(hgvs_text, edittype))
return None
try:
gene_id = GeneID(lex.gene_name)
except TypeError:
# no gene_name? it happens.
gene_id = None
print('[%s]' % hgvs_text, lex.gene_name, '(Gene ID: %s)' % gene_id)
pmid_results = {}
pmid_results['PubTator'] = PubtatorHgvs2Pmid(lex)
pmid_results['ClinVar'] = ClinvarHgvs2Pmid(lex)
return pmid_results
def test_id(self):
gene = Gene(self.gene_id)
assert_that(GeneID(self.gene_id), is_(self.gene_id))
assert_that(GeneName(self.gene_id), is_(self.gene_name))
assert_that(gene.id, is_(self.gene_id))
assert_that(gene.name, is_(self.gene_name))