-
Notifications
You must be signed in to change notification settings - Fork 0
/
mutual_information.py
197 lines (146 loc) · 6.81 KB
/
mutual_information.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
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
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
from email.mime import base
from lib2to3.pgen2.pgen import generate_grammar
import os
import pandas as pd
import numpy as np
from sklearn import feature_selection
from sklearn import metrics
import seaborn as sns
import matplotlib.pyplot as plt
main_gene_symbol = ["OLIG2", "PDPN" ,"POSTN" ,"MECOM" ,"EZH2" ,"HIF1A" ,"BIRC5" ,"ID1" ,"ID2" ,"IGFBP2" ,"ITGA6" ,"DANCR" ,"MET" ,"MYC" ,"NOS2" ,"PDGFRA" ,"PI3" ,"TGFBR2" ,"TNFAIP3", "PROM1", "CD44"]
interesting_region = ["expression_energy_it", "expression_energy_ct", "expression_energy_le"]
def calc_MI(x, y, bins):
c_xy = np.histogram2d(x, y, bins)[0]
mi = metrics.mutual_info_score(None, None, contingency=c_xy)
return mi
def find_common_gene_expression(gene_file):
gene_expression = pd.read_csv(gene_file)
gene_expression = gene_expression.fillna(0)
patients = gene_expression.tumor_name.unique()
it_gene_df = pd.DataFrame(index=patients, columns=main_gene_symbol)
ct_gene_df = pd.DataFrame(index=patients, columns=main_gene_symbol)
le_gene_df = pd.DataFrame(index=patients, columns=main_gene_symbol)
for gene in main_gene_symbol:
for patient in patients:
df = gene_expression[gene_expression['tumor_name'] == patient]
df = df[df['gene_symbol'] == gene]
for region in interesting_region:
gene_mean = df[region].mean()
if region == "expression_energy_it":
it_gene_df.at[patient, gene] = gene_mean
if region == "expression_energy_ct":
ct_gene_df.at[patient, gene] = gene_mean
if region == "expression_energy_le":
le_gene_df.at[patient, gene] = gene_mean
for index in it_gene_df.index:
if "-2" in index:
it_gene_df = it_gene_df.drop([index])
else:
it_gene_df = it_gene_df.rename(index={index:index.split('-')[0]})
it_gene_df = it_gene_df.sort_index()
for index in ct_gene_df.index:
if "-2" in index:
ct_gene_df = ct_gene_df.drop([index])
else:
ct_gene_df = ct_gene_df.rename(index={index:index.split('-')[0]})
ct_gene_df = ct_gene_df.sort_index()
for index in le_gene_df.index:
if "-2" in index:
le_gene_df = le_gene_df.drop([index])
else:
le_gene_df = le_gene_df.rename(index={index:index.split('-')[0]})
le_gene_df = le_gene_df.sort_index()
return it_gene_df, ct_gene_df, le_gene_df
def feature_sum_up(feature_file):
features = pd.read_csv(feature_file)
patients = features.Patient.unique()
feature_col = features.columns[28:-1]
feature_df = pd.DataFrame(index=patients, columns=feature_col)
for patient in patients:
df = features[features['Patient'] == patient]
for feature in feature_col:
feature_mean = df[feature].mean()
feature_df.at[patient, feature] = feature_mean
return feature_df.sort_index()
if __name__ == '__main__':
base_path = os.path.dirname(__file__)
gene_file = os.path.join(base_path, 'data', 'gene_expression_details.csv')
it_gene_df, ct_gene_df, le_gene_df = find_common_gene_expression(gene_file)
feature_file = os.path.join(base_path, "data", "feature_extraction.csv")
feature_df_origin = feature_sum_up(feature_file)
feature_df = feature_df_origin.copy(deep=True)
mutual_info_df = pd.DataFrame(index=feature_df.columns, columns=main_gene_symbol)
# print(it_gene_df)
# print(feature_df)
# for index in it_gene_df.index:
# if index not in feature_df.index:
# it_gene_df = it_gene_df.drop([index])
# for index in feature_df.index:
# if index not in it_gene_df.index:
# feature_df = feature_df.drop([index])
for index in ct_gene_df.index:
if index not in feature_df.index:
ct_gene_df = ct_gene_df.drop([index])
for index in feature_df.index:
if index not in ct_gene_df.index:
feature_df = feature_df.drop([index])
# print(it_gene_df)
# print(feature_df)
# feature_np = feature_df.to_numpy()
MI_df = pd.DataFrame(index=feature_df.columns, columns=ct_gene_df.columns)
for col1 in feature_df.columns:
for col2 in ct_gene_df.columns:
feature_vec = feature_df[col1].to_numpy()
gene_vec = ct_gene_df[col2].to_numpy()
# print(feature_np.shape)
# print(gene_vec.shape)
# MU = feature_selection.mutual_info_classif(X=feature_np, y=gene_vec)
# print(MU)
MI = calc_MI(feature_vec, gene_vec, 5)
MI_df.at[col1, col2] = MI
# print(MI_df)
fig = plt.figure(figsize=(30, 30))
sns.heatmap(MI_df.fillna(0), annot=True, fmt="g")
heatmap_path = os.path.join(base_path, 'data', 'Mutual_Information', 'Radiomics_with_ctAreaGeneExpression.svg')
plt.savefig(heatmap_path, format='svg', dpi=1800)
plt.close(fig)
feature_df = feature_df_origin.copy(deep=True)
for index in it_gene_df.index:
if index not in feature_df.index:
it_gene_df = it_gene_df.drop([index])
for index in feature_df.index:
if index not in it_gene_df.index:
feature_df = feature_df.drop([index])
MI_df = pd.DataFrame(index=feature_df.columns, columns=it_gene_df.columns)
for col1 in feature_df.columns:
for col2 in it_gene_df.columns:
feature_vec = feature_df[col1].to_numpy()
gene_vec = it_gene_df[col2].to_numpy()
MI = calc_MI(feature_vec, gene_vec, 5)
MI_df.at[col1, col2] = MI
# print(MI_df)
fig = plt.figure(figsize=(30, 30))
sns.heatmap(MI_df.fillna(0), annot=True, fmt="g")
heatmap_path = os.path.join(base_path, 'data', 'Mutual_Information', 'Radiomics_with_itAreaGeneExpression.svg')
plt.savefig(heatmap_path, format='svg', dpi=1800)
plt.close(fig)
feature_df = feature_df_origin.copy(deep=True)
for index in le_gene_df.index:
if index not in feature_df.index:
le_gene_df = le_gene_df.drop([index])
for index in feature_df.index:
if index not in le_gene_df.index:
feature_df = feature_df.drop([index])
MI_df = pd.DataFrame(index=feature_df.columns, columns=le_gene_df.columns)
for col1 in feature_df.columns:
for col2 in le_gene_df.columns:
feature_vec = feature_df[col1].to_numpy()
gene_vec = le_gene_df[col2].to_numpy()
MI = calc_MI(feature_vec, gene_vec, 5)
MI_df.at[col1, col2] = MI
# print(MI_df)
fig = plt.figure(figsize=(30, 30))
sns.heatmap(MI_df.fillna(0), annot=True, fmt="g")
heatmap_path = os.path.join(base_path, 'data', 'Mutual_Information', 'Radiomics_with_leAreaGeneExpression.svg')
plt.savefig(heatmap_path, format='svg', dpi=1800)
plt.close(fig)