def __init__(self):
self.blw = BioLinkApiWrapper(Config().get_biolink_api_endpoint())
self.mg = get_client('gene')
self.meta = {
'taxon': 'human',
'limit': None,
}
class GeneInteractions:
def __init__(self):
self.blw = BioLinkApiWrapper(Config().get_biolink_api_endpoint())
@staticmethod
# RMB: July 5, 2019 - gene_records is a Pandas DataFrame
def load_gene_set(gene_records):
annotated_gene_set = []
for gene in gene_records.to_dict(orient='records'):
if not gene['hit_symbol']:
gene['hit_symbol'] = \
Resolver.get_the_resolver(). \
translate_one(source=gene['hit_id'], identifier_range=SYMBOL)
annotated_gene_set.append({
'input_id': gene['hit_id'],
'sim_input_curie': gene['hit_id'],
'input_symbol': gene['hit_symbol']
})
return annotated_gene_set
def get_interactions(self, input_gene_set, threshold):
annotated_input_gene_set = self.load_gene_set(input_gene_set)
lower_bound = int(threshold)
results = []
for gene in annotated_input_gene_set:
interactions = self.blw.gene_interactions(
gene_curie=gene['sim_input_curie'])
for assoc in interactions['associations']:
interaction = \
self.blw.parse_association(
input_id=gene['sim_input_curie'],
input_label=gene['input_symbol'],
association=assoc
)
results.append({
'input_id': interaction['input_id'],
'input_symbol': interaction['input_symbol'],
'hit_symbol': interaction['hit_symbol'],
'hit_id': interaction['hit_id'],
'score': 1, # CX: changed score from 0 to 1
})
# Process the results
results = pd.DataFrame().from_records(results)
counts = results['hit_symbol'].value_counts().rename_axis(
'unique_values').to_frame('counts').reset_index()
high_counts = counts[
counts['counts'] > lower_bound]['unique_values'].tolist()
results = pd.DataFrame(
results[results['hit_symbol'].isin(high_counts)])
# CX: remove results where input gene = output gene. Output gene can still be disease associated genes.
results = results[~(results['hit_symbol'] == results['input_symbol'])]
return results
class LookUp(object):
def __init__(self):
self.blw = BioLinkApiWrapper(Config().get_biolink_api_endpoint())
self.mg = get_client('gene')
self.meta = {
'taxon': 'human',
'limit': None,
}
def metadata(self):
print("""Mod O DiseaseGeneLookup metadata:""")
pprint(self.meta)
def disease_geneset_lookup(self,
disease_identifier,
disease_label,
query_biolink=True):
disease_gene_association_results = self.blw.disease2genes(
disease_identifier)
input_gene_set = [
self.blw.parse_association(disease_identifier, disease_label,
association)
for association in disease_gene_association_results['associations']
]
for input_gene in input_gene_set:
if query_biolink:
igene_mg = self.mg.query(input_gene['hit_id'].replace(
'HGNC', 'hgnc'),
species='human',
entrezonly=True,
fields='entrez,HGNC,symbol')
if igene_mg['hits']:
input_gene.update({
'input_ncbi':
'NCBIGene:{}'.format(igene_mg['hits'][0]['_id'])
})
input_genes_df = pd.DataFrame(data=input_gene_set)
if not input_genes_df.empty:
# group duplicate identifier and gather sources
input_genes_df['sources'] = input_genes_df['sources'].str.join(
', ')
input_genes_df = input_genes_df.groupby([
'input_id', 'input_symbol', 'hit_id', 'hit_symbol', 'relation'
])['sources'].apply(', '.join).reset_index()
return input_genes_df
def __init__(self):
self.blw = BioLinkApiWrapper(Config().get_biolink_api_endpoint())