-
Notifications
You must be signed in to change notification settings - Fork 0
/
clu.py
195 lines (150 loc) · 9.3 KB
/
clu.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
import pandas as pd
from scipy.stats import pearsonr
import numpy as np
import matplotlib.pyplot as plt
import scipy.cluster.hierarchy as sch
import seaborn as sns
from matplotlib.colors import LinearSegmentedColormap
from scipy.cluster.hierarchy import linkage, dendrogram
from scipy.spatial.distance import pdist
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import MinMaxScaler
import os
np.random.seed(3407)
blue_gold_cmap = LinearSegmentedColormap.from_list(
'custom_blue_gold',
['#023858', '#045a8d', '#0570b0', '#3690c0', '#74a9cf', '#a6bddb', '#d0d1e6', '#fee391', '#fec44f', '#fe9929', '#ec7014', '#cc4c02', '#993404', '#662506']
)
'''
def plot_heatmap(expression_gt, matched_spot_expression_pred, top_k=500):
# take mean of expression
mean = np.mean(expression_gt, axis=1)
# take ind of top 100
ind = np.argpartition(mean, -top_k)[-top_k:]
# Compute the correlation matrix
corr_matrix = np.corrcoef(expression_gt[ind, :])
dendrogram = sch.dendrogram(sch.linkage(corr_matrix, method='ward'), no_plot=True)
cluster_idx = dendrogram['leaves']
with open('cluster_idx_list_10x.txt', 'w') as f:
for item in cluster_idx:
f.write(f"{item}\n") # 将每个元素写入文件，每个元素占一行
cluster_idx = []
with open('cluster_idx_list_10x.txt', 'r') as f:
# 逐行读取文件内容，并添加到列表中
for line in f:
# 去除每行末尾的换行符，并转换为相应的数据类型（例如整数）
cluster_idx.append(int(line.strip()))
corr_matrix = np.corrcoef(matched_spot_expression_pred[ind, :].T)
#corr_matrix = np.corrcoef(expression_gt[ind, :])
corr_matrix_a = np.where(corr_matrix > 0.99, corr_matrix, corr_matrix - 0)
corr_matrix = corr_matrix_a[cluster_idx, :]
corr_matrix = corr_matrix[:, cluster_idx]
# Reorder the correlation matrix and plot the heatmap
plt.figure(dpi=300, figsize=(5, 5))
sns.heatmap(corr_matrix, cmap=blue_gold_cmap, xticklabels=False, yticklabels=False, cbar=True, vmin=-1, vmax=1)
#plt.colorbar()
#plt.savefig(rf"C:\DELL\Desktop\mclSTExp_attention\suppleme_Fig\heatmap/mclSTExp_10x.pdf", format="pdf", bbox_inches="tight")
plt.show()
VDiT_csv1 = pd.read_csv('D:\\PyCharm\\Py_Projects\\DiTT\\result\\data_ucec\\original.csv', index_col=0)
VDiT_csv2 = pd.read_csv('D:\\PyCharm\\Py_Projects\\DiTT\\result\\data_ucec\\prediction.csv', index_col=0)
DeepImpute_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepImpute/DeepImpute_ucec_prediction.csv', index_col=0)
AutoImpute_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/AutoImpute/AutoImpute_ucec_prediction.csv', index_col=0)
DCA_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DCA/DCA_ucec_prediction.csv', index_col=0)
CpG_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/CpG/CpG_ucec_prediction.csv', index_col=0)
scVI_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/scVI/scVI_ucec_prediction.csv', index_col=0)
DeepMicroGen_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepMicroGen/DeepMicroGen_ucec_prediction.csv', index_col=0)
data_list = [VDiT_csv1, VDiT_csv2, DeepImpute_csv2, AutoImpute_csv2, DCA_csv2, CpG_csv2, scVI_csv2, DeepMicroGen_csv2]
for data in data_list:
plot_heatmap(VDiT_csv1.values, data.values)
'''
'''
def map(dataset1, dataset2, i):
scaler = MinMaxScaler()
#scaler = StandardScaler()
#dataset1 = scaler.fit_transform(dataset1)
#dataset2 = scaler.fit_transform(dataset2)
datasets = np.corrcoef(dataset1.T)
map_ndarray = np.corrcoef(dataset2.T)
datasets = scaler.fit_transform(datasets)
map_ndarray = scaler.fit_transform(map_ndarray)
#map_ndarray[np.isnan(map_ndarray)] = 0
#datasets[np.isnan(datasets)] = 0
yellow_cmap = LinearSegmentedColormap.from_list('custom_yellow', ['lightyellow', 'gold', 'orange'])
blue_gold_cmap = LinearSegmentedColormap.from_list(
'custom_blue_gold',
['#023858', '#045a8d', '#0570b0', '#3690c0', '#74a9cf', '#a6bddb', '#d0d1e6', '#fee391', '#fec44f', '#fe9929', '#ec7014', '#cc4c02', '#993404', '#662506']
)
dynamic_cmap = LinearSegmentedColormap.from_list(
'custom_dynamic',
['#4B0082', '#9400D3', '#FF0000', '#FF4500', '#FFA500', '#FFFF00']
)
row_clusters = linkage(pdist(datasets, metric='euclidean'), method='ward') #dataset2 dataset1
col_clusters = linkage(pdist(datasets.T, metric='euclidean'), method='ward')
row_order = dendrogram(row_clusters, no_plot=True)['leaves']
col_order = dendrogram(col_clusters, no_plot=True)['leaves']
test_data = pd.DataFrame(map_ndarray)
sorted_data = test_data.iloc[row_order, col_order]
data_normalized = scaler.fit_transform(sorted_data)
ax = sns.heatmap(sorted_data, cmap=blue_gold_cmap, cbar=True, annot=False) #画图1用data_normalized 画图4用sorted_data
save_dir = 'D:\\PyCharm\\Py_Projects\\DiTT\\heatmap_4\\KRCCC\\'
os.makedirs(save_dir, exist_ok=True)
save_path = os.path.join(save_dir, 'heatmap_{:02d}.pdf'.format(i))
#plt.savefig(save_path, format='pdf', bbox_inches='tight')
plt.show()
VDiT_csv1 = pd.read_csv('D:\\PyCharm\\Py_Projects\\DiTT\\result\\data_hnsc\\original.csv', index_col=0)
VDiT_csv2 = pd.read_csv('D:\\PyCharm\\Py_Projects\\DiTT\\result\\data_hnsc\\prediction.csv', index_col=0)
DeepImpute_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepImpute/DeepImpute_hnsc_prediction.csv', index_col=0)
AutoImpute_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/AutoImpute/AutoImpute_hnsc_prediction.csv', index_col=0)
DCA_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DCA/DCA_hnsc_prediction.csv', index_col=0)
CpG_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/CpG/CpG_hnsc_prediction.csv', index_col=0)
scVI_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/scVI/scVI_hnsc_prediction.csv', index_col=0)
DeepMicroGen_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepMicroGen/DeepMicroGen_hnsc_prediction.csv', index_col=0)
data_list = [VDiT_csv1, VDiT_csv2, DeepImpute_csv2, AutoImpute_csv2, DCA_csv2, CpG_csv2, scVI_csv2, DeepMicroGen_csv2]
i = 0
for data in data_list:
map(VDiT_csv1, data, i)
i = i + 1
'''
def pcc_3(csv1, csv2):
# 获取列数（假设两个文件的列数相同）
num_columns = len(csv1.columns)
# 初始化 PCC 结果列表
pcc_results = []
# 计算每一列的 PCC
for i in range(num_columns):
pcc = csv1.iloc[:, i].corr(csv2.iloc[:, i])
pcc_results.append(pcc)
return pcc_results
VDiT_csv1 = pd.read_csv('D:\\PyCharm\\Py_Projects\\DiTT\\result\\data_hnsc\\original.csv', index_col=0)
VDiT_csv2 = pd.read_csv('D:\\PyCharm\\Py_Projects\\DiTT\\result\\data_hnsc\\prediction.csv', index_col=0)
DeepImpute_csv1 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepImpute/DeepImpute_hnsc_origin.csv', index_col=0)
DeepImpute_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepImpute/DeepImpute_hnsc_prediction.csv', index_col=0)
AutoImpute_csv1 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/AutoImpute/AutoImpute_hnsc_origin.csv', index_col=0)
AutoImpute_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/AutoImpute/AutoImpute_hnsc_prediction.csv', index_col=0)
DCA_csv1 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DCA/DCA_hnsc_origin.csv', index_col=0)
DCA_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DCA/DCA_hnsc_prediction.csv', index_col=0)
CpG_csv1 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/CpG/CpG_hnsc_origin.csv', index_col=0)
CpG_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/CpG/CpG_hnsc_prediction.csv', index_col=0)
scVI_csv1 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/CpG/CpG_hnsc_origin.csv', index_col=0)
scVI_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/scVI/scVI_hnsc_prediction.csv', index_col=0)
DeepMicroGen_csv1 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepMicroGen/DeepMicroGen_hnsc_origin.csv', index_col=0)
DeepMicroGen_csv2 = pd.read_csv('D:/PyCharm/Py_Projects/Diffusion/baseline_result/DeepMicroGen/DeepMicroGen_hnsc_prediction.csv', index_col=0)
data_list = [[VDiT_csv1, VDiT_csv2], [DeepImpute_csv1, DeepImpute_csv2], [AutoImpute_csv1, AutoImpute_csv2],
[DCA_csv1, DCA_csv2], [CpG_csv1, CpG_csv2], [DeepMicroGen_csv1, DeepMicroGen_csv2], [scVI_csv1, scVI_csv2]]
pcc_results = []
for data in data_list:
pcc_results.append(pcc_3(data[0], data[1]))
import numpy as np
t = pcc_results[0]
for i in range(1, len(pcc_results)):
t1 = np.array(pcc_results[i])
t = np.vstack((t, t1))
data_map = t.T
sorted_lists = sorted(zip(t[0], t[1], t[2], t[3], t[4], t[5], t[6]), reverse=True)
s0, s1, s2, s3, s4, s5, s6 = zip(*sorted_lists)
tuples = [s0, s1, s2, s3, s4, s5, s6]
tuples = np.array(tuples)
sns.heatmap(tuples.T, cmap=blue_gold_cmap, cbar=True, annot=False)
#sns.heatmap(data_map, cmap=blue_gold_cmap, cbar=True, annot=False)
plt.savefig('D:\TCGA\mbVDiT_data\WXS\solid\hnsc\Fig4\heatmap.pdf', format='pdf', bbox_inches='tight')
plt.show() #0:VDiT 1:DeepImpute 2:AutoImpute 3:DCA 4:DeepMicroGen 5:CpG 6:scVI