--- a
+++ b/benchmark/drug2smiles.py
@@ -0,0 +1,80 @@
+'''
+## drug maps to smiles
+## input:  "data/drugbank_drugs_info.csv"
+## output:  "data/drug2smiles.pkl"
+
+'''
+
+
+
+import csv, pickle 
+from collections import defaultdict 
+def drug2smiles_func():
+	file = "data/drugbank_drugs_info.csv"
+	with open(file, 'r') as csvfile:
+		reader = list(csv.reader(csvfile, delimiter = ','))[1:]
+	drug2smiles = defaultdict(set)
+	drug2smiles2 = dict()
+	for row in reader:
+		smiles = row[27]
+		if smiles.strip()=='':
+			continue 
+		drug1 = row[3].lower()
+		drug2 = row[11].lower()
+		drug2smiles[drug1].add(smiles)
+		drug2smiles[drug2].add(smiles)
+	for drug, smiles in drug2smiles.items():
+		smiles = list(smiles)[0]
+		drug2smiles2[drug] = smiles 
+	#### to improve 
+	'''
+		 7: 53, 3: 1452, 1: 26851, 5: 178, 10: 16, 14: 6, 2: 6129, 
+		 4: 504, 17: 8, 12: 8, 8: 38, 6: 83, 11: 12, 9: 17, 161: 1, 
+		 21: 2, 15: 4, 32: 2, 13: 2, 31: 1, 22: 2, 23: 3, 16: 1, 18: 2, 104: 1, 19: 2
+	'''
+	return drug2smiles2
+
+### disease -> icd code
+if __name__ == "__main__":
+	drug2smiles = drug2smiles_func()
+	drug2smiles_file = "data/drug2smiles.pkl"
+	pickle.dump(drug2smiles, open(drug2smiles_file, 'wb'))
+
+
+'''
+[
+ 0: 'id', 
+ 1: 'trial_id', 
+ 2: 'kind', 
+ 3: 'title', 
+ 4: 'description', 
+ 5: 'id', 
+ 6: 'intervention_id', 
+ 7: 'drug_id', 
+ 8: 'id', 
+ 9: 'type', 
+ 10: 'drugbank_id', 
+ 11: 'name', 
+ 12: 'state',   ----- solid liquid 
+ 13: 'description', 
+ 14: 'cas_number', 
+ 15: 'protein_formula', 
+ 16: 'protein_weight', 
+ 17: 'investigational', 
+ 18: 'approved', 
+ 19: 'vet_approved', 
+ 20: 'experimental', 
+ 21: 'nutraceutical', 
+ 22: 'illicit', 
+ 23: 'withdrawn', 
+ 'moldb_mono_mass', 
+ 'moldb_inchi', 
+ 'moldb_inchikey', 
+ 'moldb_smiles', 
+ 'moldb_average_mass', 
+ 'moldb_formula', 
+ 'synthesis_patent_id', 
+ 'protein_weight_details', 
+ 'biotech_kind']
+
+'''