add utils function to download HIV data and fix the tutorial 2 bugs

DeepPurpose/dataset.py

@@ -2,7 +2,7 @@
 import numpy as np
 import wget
 from zipfile import ZipFile 
-from DeepPurpose.utils import convert_y_unit
+from DeepPurpose.utils import *
 import json
 import os 
 
@@ -393,6 +393,35 @@ def load_AID1706_SARS_CoV_3CL(path = './data', binary = True, threshold = 15, ba
 	print('Done!')
 	return np.array(X_drug), target, np.array(y)
 
+def load_HIV(path = './data'):
+	download_unzip('HIV', path, 'hiv.csv')
+
+	df = pd.read_csv(os.path.join(path,'hiv.csv'))
+	df = df.iloc[df['smiles'].drop_duplicates(keep = False).index.values]
+
+	df = df[df["HIV_active"].notnull()].reset_index(drop = True)
+	y = df["HIV_active"].values
+	drugs = df.smiles.values
+	drugs_idx = np.array(list(range(len(drugs))))
+
+	return drugs, y, drugs_idx	
+
+def load_AqSolDB(path = './data'):
+
+	if os.path.exists(os.path.join(path,'curated-solubility-dataset.csv')):
+		print('Dataset already downloaded in the local system...', flush = True, file = sys.stderr)
+	else:
+		wget.download('https://dataverse.harvard.edu/api/access/datafile/3407241?format=original&gbrecs=true', path)
+
+	df = pd.read_csv(os.path.join(path,'curated-solubility-dataset.csv'))
+	df = df.iloc[df['SMILES'].drop_duplicates(keep = False).index.values]
+
+	y = df["Solubility"].values
+	drugs = df.SMILES.values
+	drugs_idx = df.Name.values
+
+	return drugs, y, drugs_idx
+
 def load_broad_repurposing_hub(path = './data'):
 	url = 'https://deeppurpose.s3.amazonaws.com/broad.csv'
 	if not os.path.exists(path):

DeepPurpose/utils.py

@@ -19,6 +19,7 @@
 import wget
 from zipfile import ZipFile 
 import os
+import sys
 
 # ESPF encoding
 vocab_path = './DeepPurpose/ESPF/drug_codes_chembl_freq_1500.txt'
@@ -871,6 +872,26 @@ def load_dict(path):
 	with open(path + '/config.pkl', 'rb') as f:
 		return pickle.load(f)
 
+URLs = {
+	'HIV': 'https://s3-us-west-1.amazonaws.com/deepchem.io/datasets/molnet_publish/hiv.zip'
+	}
+
+
+def download_unzip(name, path, file_name):
+	if not os.path.exists(path):
+		os.mkdir(path)
+
+	if os.path.exists(os.path.join(path, file_name)):
+		print('Dataset already downloaded in the local system...', flush = True, file = sys.stderr)
+	else:
+		print('Download zip file...', flush = True, file = sys.stderr)
+		url = URLs[name]
+		saved_path = wget.download(url, path)
+
+		print('Extract zip file...', flush = True, file = sys.stderr)
+		with ZipFile(saved_path, 'r') as zip: 
+		    zip.extractall(path = path) 
+
 def download_pretrained_model(model_name, save_dir = './save_folder'):
 	if model_name == 'DeepDTA_DAVIS':
 		print('Beginning Downloading DeepDTA_DAVIS Model...')