-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathget_results_test_set.py
More file actions
151 lines (130 loc) · 7.98 KB
/
Copy pathget_results_test_set.py
File metadata and controls
151 lines (130 loc) · 7.98 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
import os
import pandas as pd
from deepmol.pipeline import Pipeline, VotingPipeline
from deepmol.datasets import SmilesDataset
from deepmol_models.utils.tdc_benchmark_groups import get_benchmark_group
from tdc.benchmark_group import admet_group
def get_results_for_test_set_for_specific_trial(trial, dataset_path, dataset_name, tdc_dataset_name):
pipeline_path = os.path.join(dataset_path, f'{dataset_name}', f'trial_{trial}')
pipeline = Pipeline.load(pipeline_path)
group = admet_group(path='data/')
benchmark = group.get(tdc_dataset_name)
name = benchmark['name']
predictions_list = []
for seed in [1, 2, 3, 4, 5]:
predictions = {}
train_val, test = benchmark['train_val'], benchmark['test']
train_dataset = SmilesDataset(smiles=train_val['Drug'].values, ids=train_val['Drug_ID'].values,
y=train_val['Y'].values)
test_set = SmilesDataset(smiles=test['Drug'].values, ids=test['Drug_ID'].values,
y=test['Y'].values)
pipeline.fit(train_dataset)
try:
y_pred_test = pipeline.predict_proba(test_set)
if len(y_pred_test.shape) > 1:
y_pred_test = y_pred_test[:, 1]
except:
y_pred_test = pipeline.predict(test_set)
predictions[name] = y_pred_test
predictions_list.append(predictions)
results = group.evaluate_many(predictions_list)
for _, value in results.items():
print(f'Average: {value[0]}, Std: {value[1]}')
def get_results_for_test_set(dataset_path, dataset_name, tdc_dataset_name, objective):
"""
This function is used to get results for the test set.
:param dataset_path: path to the dataset
:return: None
"""
results = os.path.join(dataset_path, f'{dataset_name}_trials.csv')
results = pd.read_csv(results)
if objective == 'minimize':
results = results.sort_values(by='value', ascending=True)
else:
results = results.sort_values(by='value', ascending=False)
results = results.head(10)
if not os.path.exists(os.path.join(dataset_path, f'{dataset_name}_test_set.csv')):
results_dataframes = pd.DataFrame(columns=['Trial', 'Average', 'Std'])
else:
results_dataframes = pd.read_csv(os.path.join(dataset_path, f'{dataset_name}_test_set.csv'))
group = admet_group(path='data/')
benchmark = group.get(tdc_dataset_name)
print("Starting for dataset: ", dataset_name)
print(f'Pipeline: {results.iloc[0]["number"]}, metric: {results.iloc[0]["value"]}')
for i, row in results.iterrows():
number = row["number"]
value = row["value"]
print(f'Pipeline: {number}, metric: {value}')
pipeline_path = os.path.join(dataset_path, f'{dataset_name}', f'trial_{number}')
pipeline = Pipeline.load(pipeline_path)
name = benchmark['name']
predictions_list = []
for seed in [1, 2, 3, 4, 5]:
predictions = {}
train_val, test = benchmark['train_val'], benchmark['test']
train_dataset = SmilesDataset(smiles=train_val['Drug'].values, ids=train_val['Drug_ID'].values,
y=train_val['Y'].values)
test_set = SmilesDataset(smiles=test['Drug'].values, ids=test['Drug_ID'].values,
y=test['Y'].values)
pipeline.fit(train_dataset)
try:
y_pred_test = pipeline.predict_proba(test_set)
if len(y_pred_test.shape) > 1:
y_pred_test = y_pred_test[:, 1]
except:
y_pred_test = pipeline.predict(test_set)
predictions[name] = y_pred_test
predictions_list.append(predictions)
results = group.evaluate_many(predictions_list)
for _, value in results.items():
results_dataframes.loc[len(results_dataframes)] = [number, value[0], value[1]]
print(f'Average: {value[0]}, Std: {value[1]}')
results_dataframes.to_csv(os.path.join(dataset_path, f'{dataset_name}_test_set.csv'), index=False)
voting_pipeline_path = os.path.join(dataset_path, f'{dataset_name}', f'voting_pipeline')
voting_pipeline = VotingPipeline.load(voting_pipeline_path)
name = benchmark['name']
predictions_list = []
for seed in [1, 2, 3, 4, 5]:
predictions = {}
train_val, test = benchmark['train_val'], benchmark['test']
train_dataset = SmilesDataset(smiles=train_val['Drug'].values, ids=train_val['Drug_ID'].values,
y=train_val['Y'].values)
test_set = SmilesDataset(smiles=test['Drug'].values, ids=test['Drug_ID'].values,
y=test['Y'].values)
voting_pipeline.fit(train_dataset)
try:
y_pred_test = voting_pipeline.predict_proba(test_set)
if len(y_pred_test.shape) > 1:
y_pred_test = y_pred_test[:, 1]
except:
y_pred_test = voting_pipeline.predict(test_set)
predictions[name] = y_pred_test
predictions_list.append(predictions)
results = group.evaluate_many(predictions_list)
for _, value in results.items():
results_dataframes.loc[len(results_dataframes)] = ['voting_pipeline', value[0], value[1]]
print(f'Average: {value[0]}, Std: {value[1]}')
results_dataframes.to_csv(os.path.join(dataset_path, f'{dataset_name}_test_set.csv'), index=False)
def test_download_files():
from deepmol_models._utils import download_models
download_models()
if __name__ == '__main__':
test_download_files()
#get_results_for_test_set_for_specific_trial(74, "bioavailability", "bioavailability", tdc_dataset_name='Bioavailability_Ma')
#get_results_for_test_set(dataset_path='cyp3a4_substrate/', dataset_name='cyp3a4_substrate', tdc_dataset_name='CYP3A4_Substrate_CarbonMangels', objective='maximize')
#get_results_for_test_set(dataset_path='half_life/', dataset_name='half_life', tdc_dataset_name='Half_Life_Obach', objective='maximize')
#get_results_for_test_set(dataset_path='hia/', dataset_name='hia', tdc_dataset_name='HIA_Hou', objective='maximize')
#get_results_for_test_set(dataset_path='dili/', dataset_name='dili', tdc_dataset_name='DILI', objective='maximize')
#get_results_for_test_set(dataset_path='herg/', dataset_name='herg', tdc_dataset_name='hERG', objective='maximize')
# get_results_for_test_set(dataset_path='ld50/', dataset_name='ld50', tdc_dataset_name='LD50_Zhu', objective='minimize')
# get_results_for_test_set(dataset_path='lipophilicity/', dataset_name='lipophilicity_stacking_regressor', tdc_dataset_name='Lipophilicity_AstraZeneca', objective='minimize')
# get_results_for_test_set(dataset_path='pgp/', dataset_name='pgp', tdc_dataset_name='Pgp_Broccatelli', objective='maximize')
# get_results_for_test_set(dataset_path='ppbr/', dataset_name='ppbr', tdc_dataset_name='PPBR_AZ', objective='minimize')
# get_results_for_test_set(dataset_path='solubility/', dataset_name='solubility', tdc_dataset_name='Solubility_AqSolDB', objective='minimize')
# get_results_for_test_set(dataset_path='vdss/', dataset_name='vdss', tdc_dataset_name='VDss_Lombardo', objective='maximize')
# get_results_for_test_set(dataset_path='ames/', dataset_name='ames', tdc_dataset_name='AMES', objective='maximize')
# get_results_for_test_set(dataset_path='bbb/', dataset_name='bbb', tdc_dataset_name='BBB_Martins', objective='maximize')
# get_results_for_test_set(dataset_path='bioavailability/', dataset_name='bioavailability', tdc_dataset_name='Bioavailability_Ma', objective='maximize')
#get_results_for_test_set(dataset_path='caco/', dataset_name='caco', tdc_dataset_name='Caco2_Wang', objective='minimize')
# get_results_for_test_set(dataset_path='clearance_hepatocyte/', dataset_name='clearance_hepatocyte', tdc_dataset_name='Clearance_Hepatocyte_AZ', objective='maximize')
# get_results_for_test_set_for_specific_trial(78, "bbb", "bbb", tdc_dataset_name='BBB_Martins')