def get_clinvar(self):
eu=eutils.EUtils()
# Filters: (Pathogenic or likely pathogenic) and (frameshif, missense or nonsense)
id,args=eu.esearch({'db':'clinvar', 'term':'(clinsig+pathogenic[Filter] OR clinsig+likely+path[Filter]) AND (mol+cons+frameshift[Filter] OR mol+cons+missense[Filter] OR mol+cons+nonsense[Filter])'})
# This gives ~28000 records (variants)
# id,args=eu.esearch({'db':'clinvar', 'term':'(clinsig+pathogenic[Filter] OR clinsig+likely+path[Filter]) AND (mol+cons+frameshift[Filter] OR mol+cons+missense[Filter] OR mol+cons+nonsense[Filter]) AND (var+deletion[Filter] OR var+indel[Filter]'})
# The above query gave ~5000 records, but it missed missense mutations because of deletion and indel filters
args['db']='clinvar'
out=eu.esummary(args,count = args['count'])
x = cv.SummaryList(out)
records = x.to_list()
# Combine results into DataFrame
rows = []
for entry in records:
data = [entry['variant_id']]
data.extend(entry['variant'])
for gene in entry['genes']:
record = data[:]
record.extend(gene)
for key in entry['trait'].keys():
row = record[:]
row.extend([key,"; ".join(entry['trait'][key])])
rows.append(row)
df = pd.DataFrame(rows, columns = ['id','VariantType','VarinatName','Pathogenic','Symbol','GeneID','Trait','Source'])
df.Trait = [trait.encode('ascii','ignore') for trait in df.Trait.values]
data=[]
for k,t_v in df.groupby('GeneID'):
S1=[x for x in util.unique(t_v['VariantType']) if not pd.isnull(x)]
S2=[x[0] + "(" + x[1] + ")" for x in zip(t_v['Source'], t_v['Trait']) if not pd.isnull(x[0]) and not pd.isnull(x[1]) ]
#print S1[:], S2[:], S3[:]
data.append({'gene_id':k, 'variant_type':";".join(S1), 'source_trait':";".join(S2)})
df=pd.DataFrame(data).query('source_trait != ""')
df.to_csv(self.fn_dest_clinvar, index=False)
def fetch_pmid(self, PMID):
eu = eutils.EUtils()
args = {}
args['db'] = 'pubmed'
args['id'] = [PMID]
out = eu.efetch(args)
X = pm.FetchList(out)
X = X.to_list(['pubmed_id', 'journal', 'title', 'author'])
if len(X) == 0:
return {}
#{'journal.title': 'Pharmaceutical medicine', 'journal.day': None, 'journal.year': '2017', 'title': 'Measuring and Improving Physician Knowledge of Safety Risks Using Traditional and Online Methods in Pharmacovigilance.', 'journal.month': None, 'journal.volume': '31', 'journal.issue': '4', 'author': ['Liede A', 'Amelio J', 'Bennett J', 'Goodman H', 'Peters PM', 'Barber R', 'Kehler E', 'Michael Sprafka J'], 'pubmed_id': '28824275', 'journal.page': '257-266'}
X = X[0]
data = {
'PMID': X['pubmed_id'],
'Title': X['title'],
'Year': X['journal.year'],
'Journal': X['journal.title'],
'Month': X['journal.month'],
'Volume': X['journal.volume'],
'Issue': X['journal.issue'],
'Page': X['journal.page'],
'Authors': ", ".join(X['author'])
}
return data
def symbol(self):
return self._get_value("./Name")
def description(self):
return self._get_value("./Description")
def chromosome(self):
return self._get_value("./Chromosome")
def mim(self):
out = self._get_nodes("./Mim/int")
for i, x in enumerate(out):
if x is None: continue
out[i] = " ".join([y.text for y in x])
return out
if __name__ == "__main__":
import eutils
eu = eutils.EUtils()
out = eu.efetch({'db': 'gene', 'id': '170743,51284'})
x = FetchList(out)
print(">>>Fetch")
print(x.to_list(x.attributes() + ["summary", "type"]))
out = eu.esummary({'db': 'gene', 'id': '170743,51284,466142'})
x = SummaryList(out)
print(">>>Summary")
print(x.to_list(x.attributes()))