ChemBC/ChemBC.py at master · idrugLab/ChemBC · GitHub

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
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3' # ------> hide info
import rdkit
import deepchem as dc
from rdkit import Chem
import numpy as np
from deepchem.feat.base_classes import MolecularFeaturizer
from deepchem.utils.typing import RDKitMol
import joblib
import argparse
from sklearn.impute import SimpleImputer
from sklearn.feature_selection import SelectPercentile,VarianceThreshold
from rdkit.Chem.AtomPairs import Pairs
from tensorflow import keras
import pandas as pd
from sklearn.base import TransformerMixin

model_path = {
    'Bcap37':'./models/Bcap37.model',
    'BT-20':'./models/BT-20.model',
    'BT-474':'./models/BT-474.model',
    'BT-549':'./models/BT-549.model',
    'HS-578T':'./models/HS-578T.model',
    'MCF-7':'./models/MCF-7.model',
    'MDA-MB-231':'./models/MDA-MB-231.model',
    'MDA-MB-361':'./models/MDA-MB-361.model',
    'MDA-MB-435':'./models/MDA-MB-435.model',
    'MDA-MB-453':'./models/MDA-MB-453.model',
    'MDA-MB-468':'./models/MDA-MB-468.model',
    'SK-BR-3':'./models/SK-BR-3.model',
    'T-47D':'./models/T-47D.model',
    'HBL-100':'./models/HBL-100.model'
}


class model():

    def __init__(self,system,dataset_path):
        self.system = system
        self.dataset_path = dataset_path
        self.model_name = 'Morgan'
        # if self.system == 'BT-20' or self.system == 'HS-578T':
        #     self.model_name = 'rdkit'
        # elif self.system == 'BT-474':
        #     self.model_name = 'MACCS'
        # elif self.system == 'Bcap37':
        #     self.model_name = 'at'
        # else:
        #     self.model_name = 'Morgan'

    def load_dataset(self):# ----> load datasets
        # if self.model_name == 'MACCS':
        #     featurizer = dc.feat.MACCSKeysFingerprint()
        #     loader = dc.data.CSVLoader(tasks=[], feature_field="Smiles", featurizer=featurizer)  # smiles_field指smiles列的标签
        #     dataset_origin = loader.create_dataset(self.dataset_path, shard_size=8192)
        #     dataset = self.load(dataset_origin)
        #     return dataset
        # if self.model_name == 'rdkit':
        #     featurizer = dc.feat.RDKitDescriptors()
        #     loader = dc.data.CSVLoader(tasks=[], feature_field="Smiles", featurizer=featurizer)  # smiles_field指smiles列的标签
        #     dataset_origin = loader.create_dataset(self.dataset_path, shard_size=8192)
        #     dataset = self.feature_dataset(dataset_origin)
        #     transformer = dc.trans.MinMaxTransformer(transform_X=True, dataset=dataset)
        #     dataset = transformer.transform(dataset)
        #     return dataset
        # if self.model_name == 'at':
        #     featurizers = AtomPairFeaturizer()
        #     loader = dc.data.CSVLoader(tasks=[],feature_field="Smiles",featurizer=featurizers)
        #     dataset_origin = loader.create_dataset(self.dataset_path, shard_size=8192)
        #     dataset = self.load(dataset_origin)
        #     return dataset
        # if self.model_name == 'Morgan':
        featurizer = dc.feat.CircularFingerprint(size=1024)
        loader = dc.data.CSVLoader(tasks=[], feature_field="Smiles", featurizer=featurizer)  # smiles_field指smiles列的标签
        dataset_origin = loader.create_dataset(self.dataset_path, shard_size=8192)
        dataset = self.load(dataset_origin)
        return dataset


    def load_model(self):# ----> load model by joblib
        mp = model_path[self.system]
        #if self.system == 'SK-BR-3':
        #   reloaded = keras.models.load_model(mp)
        #    model = dc.models.MultitaskClassifier(1, 1024, layer_sizes=[1024, 1024], weight_decay_penalty=0.001,model_dir="s")
        #    model.model = reloaded
        #    return model

        #else:
        model = joblib.load(mp)
        return model

    def run(self):# ---- >predict scores
        model = self.load_model()
        datasets = self.load_dataset()
        print('模型已加载完毕，细胞系为：%s,文件名为：%s,输出文件名为：%s_result.csv'%(self.system,self.dataset_path,self.system))
        print('开始计算……')
        y_pred = model.predict(datasets)

        y_pre = y_pred[:, 1]
        data = pd.read_csv(self.dataset_path)
        data['Score'] = y_pre
        d = pd.DataFrame(data,index=None)
        d = d.sort_values(by="Score",ascending=False)
        d.to_csv('%s_result.csv'%self.system,index=None)
        print('计算完毕！')

    def load(self,dataset):
        x_load, y_load = [],[]
        for x, _, _, id in dataset.itersamples():
            x_load.append(x)
        dataset_new = dc.data.NumpyDataset(X=x_load)
        return dataset_new

    def feature_dataset(self,dataset):
        imp = SimpleImputer(strategy = "constant",fill_value=0)
        #imp =SimpleImputer(missing_values=np.nan, strategy='mean')

        sel = VarianceThreshold(threshold=(.7 * (1 - .7)))
        selector = SelectPercentile(percentile=30)
        x_load, y_load = [],[]
        import random
        for x, y, _, id in dataset.itersamples():
            x[x == np.inf] = np.nan
            x_load.append(x)
            y_load.append(int(random.randint(0,1)))
        x_load_1 = imp.fit_transform(x_load)
        x_load_2 = sel.fit_transform(x_load_1)
        x_load_3 = selector.fit_transform(x_load_2,y_load)
        y_load = np.array(y_load)[:, np.newaxis]
        dataset_new = dc.data.NumpyDataset(X=x_load_3)
        return dataset_new

class AtomPairFeaturizer(MolecularFeaturizer):
    def _featurize(self, mol: RDKitMol) -> np.ndarray:
        fp = list(Pairs.GetHashedAtomPairFingerprint(mol))
        fp = np.asarray(fp, dtype=float)
        return fp

def main(system,files):# ----> main process
    run = model(system,files)
    run.run()

if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--files', type=str,default=None,
                        help='csv files with a coloum of smiles format')
    parser.add_argument('--system', type=str,default=None,
                        help='provide system of cells : Bcap37、BT-20、BT-474、BT-549、HS-578T、MCF-7、MDA-MB-231、MDA-MB-361、MDA-MB-435、MDA-MB-453、MDA-MB-468、SK-BR-3、T-47D、HBL-100')
    # parser.add_argument('--model', type=str,default=None,
    #                     help='provide models : RF_AtomPairs_H、RF_rdkit_H、RF_MACCS_H、RF_Morgan、RF_rdkit、DNN_Morgan')
    parser.add_argument('--all_systems', type=bool,default=False,
                        help='Choose all systems：defaul false')
    args = parser.parse_args()
    # try:
    #     main(args.files,args.system)

    # except:
    #     print('没有选择文件或者文件格式有错误。')
    all = args.all_systems
    if all:
        systems = ['Bcap37','BT-20','BT-474','BT-549','HS-578T','MCF-7','MDA-MB-231','MDA-MB-361','MDA-MB-435','MDA-MB-453','MDA-MB-468','SK-BR-3','T-47D','HBL-100']
        for s in systems:
            main(s,args.files)
            print()
    else:
        main(args.system,args.files)