def get_disease_specific_drugs(parser, selected_drugs, phenotypes):
import text_utilities
disease_to_drugs = {}
indication_to_diseases = {}
for drug, indication in parser.drug_to_indication.iteritems():
if drug not in selected_drugs:
continue
if indication is None:
continue
#if any(map(lambda x: x is not None, [ exp.search(indication) for exp in exps ])):
#disease = keywords[0]
#disease_to_drugs.setdefault(disease, set()).add(drug)
#for disease, exp in zip(phenotypes, exps):
# if exp.search(indication.lower()) is not None:
# disease_to_drugs.setdefault(disease, set()).add(drug)
indication = indication.lower()
for disease in phenotypes:
#if all([ indication.find(word.strip()) != -1 for word in disease.split(",") ]):
# disease_to_drugs.setdefault(disease, set()).add(drug)
values = text_utilities.tokenize_disease_name(disease)
#print disease, values
indication_to_diseases.setdefault(indication, set())
if all([indication.find(word.strip()) != -1 for word in values]):
#print disease, drug
disease_to_drugs.setdefault(disease, set()).add(drug)
indication_to_diseases.setdefault(indication,
set()).add(disease)
else:
values = text_utilities.tokenize_disease_name(
disease.replace("2", "II"))
if all(
[indication.find(word.strip()) != -1 for word in values]):
disease_to_drugs.setdefault(disease, set()).add(drug)
indication_to_diseases.setdefault(indication,
set()).add(disease)
else:
values = text_utilities.tokenize_disease_name(
disease.replace("1", "I"))
if all([
indication.find(word.strip()) != -1
for word in values
]):
disease_to_drugs.setdefault(disease, set()).add(drug)
indication_to_diseases.setdefault(indication,
set()).add(disease)
# Print non-matching indications #!
for indication, diseases in indication_to_diseases.iteritems():
if len(diseases) == 0:
continue
print indication.encode('ascii', 'ignore')
elif indication.find(" not ") != -1 or indication.find(
" except ") != -1:
continue
print diseases, indication.encode('ascii', 'ignore')
#print disease_to_drugs["diabetes mellitus, type 2"]
return disease_to_drugs
def get_disease_specific_drugs(parser, selected_drugs, phenotypes):
import text_utilities
disease_to_drugs = {}
indication_to_diseases = {}
for drug, indication in parser.drug_to_indication.iteritems():
if drug not in selected_drugs:
continue
if indication is None:
continue
#if any(map(lambda x: x is not None, [ exp.search(indication) for exp in exps ])):
#disease = keywords[0]
#disease_to_drugs.setdefault(disease, set()).add(drug)
#for disease, exp in zip(phenotypes, exps):
# if exp.search(indication.lower()) is not None:
# disease_to_drugs.setdefault(disease, set()).add(drug)
indication = indication.lower()
for disease in phenotypes:
#if all([ indication.find(word.strip()) != -1 for word in disease.split(",") ]):
# disease_to_drugs.setdefault(disease, set()).add(drug)
values = text_utilities.tokenize_disease_name(disease)
#print disease, values
indication_to_diseases.setdefault(indication, set())
if all([ indication.find(word.strip()) != -1 for word in values ]):
#print disease, drug
disease_to_drugs.setdefault(disease, set()).add(drug)
indication_to_diseases.setdefault(indication, set()).add(disease)
else:
values = text_utilities.tokenize_disease_name(disease.replace("2", "II"))
if all([ indication.find(word.strip()) != -1 for word in values ]):
disease_to_drugs.setdefault(disease, set()).add(drug)
indication_to_diseases.setdefault(indication, set()).add(disease)
else:
values = text_utilities.tokenize_disease_name(disease.replace("1", "I"))
if all([ indication.find(word.strip()) != -1 for word in values ]):
disease_to_drugs.setdefault(disease, set()).add(drug)
indication_to_diseases.setdefault(indication, set()).add(disease)
# Print non-matching indications #!
for indication, diseases in indication_to_diseases.iteritems():
if len(diseases) == 0:
continue
print indication.encode('ascii','ignore')
elif indication.find(" not ") != -1 or indication.find(" except ") != -1:
continue
print diseases, indication.encode('ascii','ignore')
#print disease_to_drugs["diabetes mellitus, type 2"]
return disease_to_drugs
def convert_fda_name_to_mesh(disease, mesh_name_to_ids):
phenotype = None
disease = disease.replace("^s", "").replace("'s","")
if disease in mesh_name_to_ids:
phenotype = disease
return phenotype
# Get words skipping disease / disorder / syndrome / plural / 's
values = text_utilities.tokenize_disease_name(disease, exact=False)
val_and_phenotypes = []
for mesh_name in mesh_name_to_ids:
val = sum([ mesh_name.lower().find(word.strip()) != -1 for word in values ])
#print mesh_name, val
if val > len(values) / 2.0:
#print mesh_name, disease
val_and_phenotypes.append((float(val)/len(mesh_name.split()), mesh_name))
#print values, val_and_phenotypes
if len(val_and_phenotypes) > 0:
val_and_phenotypes.sort()
phenotype = val_and_phenotypes[-1][1]
return phenotype
def convert_fda_name_to_mesh(disease, mesh_name_to_ids):
phenotype = None
disease = disease.replace("^s", "").replace("'s", "")
if disease in mesh_name_to_ids:
phenotype = disease
return phenotype
# Get words skipping disease / disorder / syndrome / plural / 's
values = text_utilities.tokenize_disease_name(disease, exact=False)
val_and_phenotypes = []
for mesh_name in mesh_name_to_ids:
val = sum(
[mesh_name.lower().find(word.strip()) != -1 for word in values])
#print mesh_name, val
if val > len(values) / 2.0:
#print mesh_name, disease
val_and_phenotypes.append(
(float(val) / len(mesh_name.split()), mesh_name))
#print values, val_and_phenotypes
if len(val_and_phenotypes) > 0:
val_and_phenotypes.sort()
phenotype = val_and_phenotypes[-1][1]
return phenotype