-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdc_compitition.py
133 lines (115 loc) · 4.28 KB
/
dc_compitition.py
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
import pandas as pd
import re
import seaborn as sns
import matplotlib.pyplot as plt
from gensim.models import Word2Vec
from sklearn.feature_selection import SelectFromModel
from sklearn.model_selection import train_test_split
import lightgbm as lgb
df_protein_train = pd.read_csv('./data/df_protein_train.csv')
df_protein_test = pd.read_csv('./data/df_protein_test.csv')
df_protein = pd.concat([df_protein_train, df_protein_test])
texts = [[word for word in re.findall(r'.{3}', doc)]
for doc in list(df_protein['Sequence'])]
n = 128
model = Word2Vec(texts, size=n, window=4, min_count=1,negative=3,
sg=1, sample=0.001, hs=1, workers=4)
vectors = pd.DataFrame([model[word] for word in (model.wv.vocab)])
vectors['Word'] = list[model.wv.vocab)
vectors.columns = ["vec_{0}".format(i) for i in range(0,n)] + ['Word']
wide_vec = pd.DataFrame()
result1 = []
aa = list(df_protein['Protein_ID'])
for i in range(len(texts)):
result2=[]
for j in range(len(texts[i])):
result2.append(aa[i])
result1.extend(result2)
wide_vec['Id'] = result1
result1 = []
for i in range(len(texts)):
result2=[]
for j in range(len(texts[i])):
result2.append(texts[i][j])
result1.extend(result2)
wide_vec['Word'] = result1
del result1,result2
wide_vec = wide_vec.merge(vectors,on='Word',how='left')
wide_vec = wide_vec.drop('Word',axis=1)
cols = ['vec_{0}'.format(i) for i in range(0,n)]
wide_vec.columns = ['Protein_ID'] + cols
del vectors
df_protein = pd.DataFrame(wide_vec.groupby(['Protein_ID'])[cols].agg('mean'))
df_protein = df_protein.reset_index()
del wide_vec
df_molecule = pd.read_csv('./data/df_molecule.csv')
feat = []
for i in range(0,len(df_molecule)):
feat.append(df_molecule['Fingerprint'][i].split(','))
feat = pd.DataFrame(feat)
feat = feat.astype('int')
feat.columns = ["Fingerprint_{0}".format(i) for i in range(0,167)]
feat['Molecule_ID'] = df_molecule['Molecule_ID']
df_molecule = df_molecule.drop('Fingerprint',axis=1)
df_molecule = df_molecule.merge(feat, on='Molecule_ID')
del feat
df_affinity_train = pd.read_csv('./data/df_affinity_train.csv')
df_affinity_predict = pd.read_csv('./data/df_affinity_test_toBePredicted.csv')
df_train = pd.merge(df_affinity_train,df_protein,on='Protein_ID',how='left')
df_train = pd.merge(df_train, df_molecule, on='Molecule_ID',how='left')
df_predict = pd.merge(df_affinity_predict,df_protein,on='Protein_ID',how='left')
df_predict = pd.merge(df_predict, df_molecule, on='Molecule_ID',how='left')
df_predict = df_predict.drop(columns=['Protein_ID','Molecule_ID'])
df_predict = df_predict.fillna(0)
del df_affinity_train
X = df_train.drop(columns=['Ki','Protein_ID','Molecule_ID']).fillna(0)
y = df_train['Ki'].fillna(0)
'''from sklearn.ensemble import RandomForestRegressor
clf = RandomForestRegressor().fit(X,y)
print(X.shape)
model = SelectFromModel(clf, prefit=True)
X = model.transform(X)
df_predict = model.transform(df_predict)
print(X.shape)
print(df_predict.shape)'''
X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.25,random_state=33)
train = lgb.Dataset(X_train, label=y_train)
test = lgb.Dataset(X_test, label=y_test, reference=train)
params={
'boosting_type' : 'gbdt',
'objective' : 'regression_l2',
'metric' : 'l2',
'min_child_weight' : 3,
'max_bin': 510,
'num_leaves' : 2**8,
#'num_leaves' : 2**5,
'lambda_l2' : 10,
'subsample' : 0.7,
'colsample_bytree' : 0.7,
'colsample_bylevel' : 0.7,
'learning_rate': 0.01,
#'learning_rate' : 0.05,
'tree_method' : 'exact',
'seed' : 2017,
'nthread' : 12,
'silent' : True
}
num_round = 5000
#num_round = 3000
gbm = lgb.train(params,
train,
num_round,
verbose_eval=50,
valid_sets=[train,test]
)
preds_sub = gbm.predict(df_predict)
df_affinity_predict['Ki'] = preds_sub
df_affinity_predict.to_csv('./lgb.csv',index=False)
#cyp_3a4, cyp_2c9, cyp_2d6, ames_toxicity, fathead_minnow_toxicity
#tetrahymena_pyriformis_toxicity, honey_bee, cell_permeability,
#logP, renal_organic_cation_transporter, CLtotal, hia, biodegradation,
#Vdd, p_glycoprotein_inhibition, NOAEL, solubility, bbb
#g = sns.FacetGrid(df_molecule)
#g.map(sns.kdeplot, 'cyp_2c9',shade=True)
#sns.pairplot(df_molecule[['cyp_3a4','cyp_2c9','cyp_2d6','ames_toxicity']].dropna())
#plt.show()