/
_Clustering_DMR_clustermap.py
340 lines (289 loc) · 17.2 KB
/
_Clustering_DMR_clustermap.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
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
# source activate complexheatmap
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib
import matplotlib.pyplot as plt
from matplotlib.patches import Patch
import scanpy as sc
import scipy.stats as stats
from statsmodels.stats.multitest import multipletests
sc.settings.verbosity = 3
plt.rcParams['figure.figsize'] = (5,5)
cmap = matplotlib.colors.LinearSegmentedColormap.from_list("", ["#104e8b", "#ffdab9", "#8b0a50"])
cmap2 = matplotlib.colors.LinearSegmentedColormap.from_list("", ["#0057b7", "#000000", "#ffd700"])
cmap3 = matplotlib.colors.LinearSegmentedColormap.from_list("", ["#0057B8", "#000000", "#ffd700"])
%matplotlib
%autoindent
# Step 1. Create Tumor Normal imputed matrix
import pandas as pd
dmr = pd.read_csv("DMR_mat_imputed_corrected.csv", index_col=0)
dmr.iloc[84:, :-1].to_csv("DMR_mat_imputed_corrected_tumor.csv")
dmr.iloc[:84, :-1].to_csv("DMR_mat_imputed_corrected_normal.csv")
# Step 2-1.
dmr_t = sc.AnnData(pd.read_csv("DMR_mat_imputed_corrected_tumor.csv", index_col=0))
dmr_t.raw = dmr_t
dmr_t.layers['Percent_met'] = dmr_t.X
clinic_info = pd.read_csv('/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/2022_WC300_clinical_information_Xadded.csv', index_col=0)
clinic_info[['T.stage', 'N.stage', 'M.stage', 'TNM.stage', 'HER2IHC']] = clinic_info[['T.stage', 'N.stage', 'M.stage', 'TNM.stage', 'HER2IHC']].astype(str)
#clinic_info = clinic_info.iloc[:,np.r_[0:17, -1]]
dmr_t.obs = clinic_info
sc.pp.scale(dmr_t)
sc.tl.pca(dmr_t, n_comps=83, zero_center=True)
#sc.pl.pca(dmr_t, color='Lauren')
# sc.pl.pca_variance_ratio(dmr_t, log=True)
pca_variance_t = pd.DataFrame(dmr_t.uns['pca']['variance_ratio'], index=list(map(lambda x: 'PC' + str(x), list(range(1,84)))), columns=['Variance_ratio'])
np.sum(pca_variance_t.values.flatten()[:13])
# 0.7076739
#sc.pp.neighbors(dmr_t, n_neighbors=11, n_pcs=13) ###
#sc.pp.neighbors(dmr_t, n_neighbors=20, n_pcs=13)
sc.pp.neighbors(dmr_t, n_neighbors=11, n_pcs=13)
sc.tl.leiden(dmr_t, resolution=1.0, key_added='leiden_r1')
sc.tl.leiden(dmr_t, resolution=0.5, key_added='leiden_r05')
sc.tl.umap(dmr_t, min_dist=0.5, spread=1.0, n_components=2, alpha=1.0, gamma=1.0, init_pos='spectral', method='umap')
sc.pl.umap(dmr_t, color=['leiden_r1', 'leiden_r05', 'EpiBurden'], add_outline=False, color_map=cmap)
leiden_name_change_dict1 = {'3': 'leiden_A',
'1': 'leiden_B',
'0': 'leiden_C',
'2': 'leiden_D'}
dmr_t.obs['DMR_leiden'] = dmr_t.obs['leiden_r05'].map(lambda x: leiden_name_change_dict1[x]).astype('category')
sc.tl.diffmap(dmr_t)
sc.pl.diffmap(dmr_t, color=['DMR_leiden'], add_outline=False, color_map=cmap)
start_cell = np.isin(dmr_t.obs['DMR_leiden'], 'leiden_A') # leiden_r1의 3번 cluster
max_start_id = np.argmax(dmr_t.obsm['X_diffmap'][start_cell, 1])
root_id = np.arange(len(start_cell))[start_cell][max_start_id]
dmr_t.uns['iroot'] = root_id
sc.tl.dpt(dmr_t, n_branchings=0, n_dcs=15)
sns.lmplot(data=dmr_t.obs, x='dpt_pseudotime', y='EpiBurden')
lin = tuple(sorted(list(dmr_t.obs['DMR_leiden'].values.unique())))
dmr_t.obs['DMR_leiden'] = dmr_t.obs['DMR_leiden'].cat.reorder_categories(list(lin), ordered=True)
color_dict = {
"leiden_A": "#d62728",
"leiden_B": "#ff7f0e",
"leiden_C": "#1f77b4",
"leiden_D": "#2ca02c"
} # equivalent to dict(zip(list(dmr_t.obs['DMR_leiden'].value_counts().index), dmr_t.uns['DMR_leiden_colors']))
sns.boxplot(data=dmr_t.obs, x='DMR_leiden', y='EpiBurden', palette=color_dict)
sns.swarmplot(data=dmr_t.obs, x='DMR_leiden', y='EpiBurden', color=".2")
rna = pd.read_table("/mnt/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/02.RNA-seq/STAD_SNUH_vst.txt", index_col=0, sep=' ')
rna = pd.read_table("/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/02.RNA-seq/GENCODE_V24/STAD_SNUH_vst.txt", index_col=0, sep=' ')
rna.columns = list(map(lambda x: "X" + x, rna.columns))
deg_tn_protein = pd.read_table("/mnt/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/02.RNA-seq/STAD_SNUH_Tumor_leiden_vst_DEG_Leiden_A_D_protein.txt", index_col="ID")
deg_tn_protein.columns = list(map(lambda x: "X" + x, deg_tn_protein.columns))
pro_met = pd.read_table("Promoter_up500down500_ALL.txt", index_col="ID")
pro_met = pd.read_table("/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/03.WGBS/Promoter_cCRE_ALL.txt", index_col="ID")
pro_met = pd.read_table("/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/03.WGBS/NEW/Promoter_cCRE_ALL.txt", index_col="ID")
pro_met.columns = list(map(lambda x: "X" + x, pro_met.columns))
pro_met_info = pd.DataFrame(list(zip(list(map(lambda x: x.split('/')[0], pro_met.index)), list(map(lambda x: x.split('/')[1], pro_met.index)), list(map(lambda x: x.split('/')[2], pro_met.index)), list(map(lambda x: x.split('/')[-1], pro_met.index)))), columns=['Loc', 'GeneID', 'EnsemblID', 'CpG'], index=pro_met.index)
pro_met.loc[pro_met_info[pro_met_info['GeneID'] == 'OR4F5'].index]
deg_up = list(map(lambda x: x.strip('\n'), open("/mnt/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/02.RNA-seq/deg_up_protein.txt", 'r').readlines()))
deg_down = list(map(lambda x: x.strip('\n'), open("/mnt/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/02.RNA-seq/deg_down_protein.txt", 'r').readlines()))
tfs = list(map(lambda x: x.strip('\n'), open("/mnt/data/Projects/phenomata/01.Projects/08.StomachCancer_backup/02.RNA-seq/hs_hgnc_curated_tfs.txt", 'r').readlines()))
tfs_in_deg =
deg_rowcolors1 = list(map(lambda x: '#800000' if x in deg_up else '#000080', list(deg_tn_protein.loc[(deg_tn_protein.index.isin(deg_up)) | (deg_tn_protein.index.isin(deg_down))].index)))
deg_rowcolors2 = list(map(lambda x: '#000000' if x in tfs else '#FFFFFF', list(deg_tn_protein.loc[(deg_tn_protein.index.isin(deg_up)) | (deg_tn_protein.index.isin(deg_down))].index)))
sns.clustermap(
rna.loc[(rna.index.isin(deg_up)) | (rna.index.isin(deg_down))].iloc[:,84:],
method="ward",
metric="euclidean",
z_score=None,
standard_scale=0,
row_cluster=True,
col_cluster=True,
xticklabels=True,
yticklabels=False,
col_colors=[col_colors1],
row_colors=deg_rowcolors1,
cmap=cmap2
)
deg_tn_protein.loc[:, (deg_tn_protein.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_A'].index)) | (deg_tn_protein.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_D'].index)) ].loc[deg_tn_protein.index.isin(deg_up)]
#rfh = open("DMR_tumor_pseudotime_and_DEG_TN_protein_Correlation.txt", 'w')
#rfh.write("ID\tPearsonR\tPvalue\n")
r_list, p_list = list(), list()
for i in range(len(deg_tn_protein.index)):
r, p = stats.pearsonr(dmr_t.obs['dpt_pseudotime'], deg_tn_protein.iloc[i])
r_list.append(r)
p_list.append(p)
multipletests(pvals=p_list, alpha=0.05, method='fdr_bh')[1]
pt_degtn = pd.DataFrame([r_list, multipletests(pvals=p_list, alpha=0.05, method='fdr_bh')[1]], columns=deg_tn_protein.index, index=['PearsonR', 'Padj']).T
leiden_name_change_dict1 = {'3': 'leiden_A',
'1': 'leiden_B',
'0': 'leiden_C',
'2': 'leiden_D'}
dmr_t.obs['DMR_leiden'] = dmr_t.obs['leiden_r05'].map(lambda x: leiden_name_change_dict1[x]).astype('category')
col_colors1 = list(dict(zip(list(dmr_t.obs['DMR_leiden'].value_counts().index), dmr_t.uns['DMR_leiden_colors']))[x] for x in dmr_t.obs['DMR_leiden'])
sample_cluster_color = pd.concat([dmr_t.obs[['DMR_leiden']], pd.DataFrame(col_colors1, index=dmr_t.obs.index, columns=['Colors'])], axis=1)
#col_colors1 = list(dict(zip(list(dmr_t.obs['DMR_leiden'].value_counts().index), dmr_t.uns['DMR_leiden_colors']))[x] for x in dmr_t.obs['DMR_leiden']) * 2
dmr_t_met = pd.DataFrame(dmr_t.raw.X.T, index=dmr_t.var.index, columns=dmr_t.obs.index)
g = sns.clustermap(dmr_t_met, method='ward', metric='euclidean', z_score=None, standard_scale=None, cmap=cmap, xticklabels=True, yticklabels=False, col_colors=[col_colors1])
deg_up_sig_cor = list()
for i in deg_up:
try:
pearson = stats.pearsonr(np.delete(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0], np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))), np.delete(deg_tn_protein.loc[i].values, np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))))
if pearson[1] < 0.05: # p value만 가지고 해야 함 나중에는
if pearson[0] < 0:
deg_up_sig_cor.append(i)
print(i + '\t' + str(pearson[0]) + '\t' + str(pearson[1]))
except (IndexError, ValueError):
pass
deg_up_sig_cor = list()
for i in deg_up:
try:
spearman = stats.spearmanr(np.delete(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0], np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))), np.delete(deg_tn_protein.loc[i].values, np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))))
if spearman[1] < 0.05:
if spearman[0] < 0: # p value만 가지고 해야 함 나중에는
deg_up_sig_cor.append(i)
print(i + '\t' + str(spearman[0]) + '\t' + str(spearman[1]))
except (IndexError, ValueError):
pass
deg_down_sig_cor = list()
for i in deg_down:
try:
pearson = stats.pearsonr(np.delete(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0], np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))), np.delete(deg_tn_protein.loc[i].values, np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))))
if pearson[1] < 0.05:
if pearson[0] < 0: # p value만 가지고 해야 함 나중에는
deg_down_sig_cor.append(i)
print(i + '\t' + str(pearson[0]) + '\t' + str(pearson[1]))
except (IndexError, ValueError):
pass
deg_down_sig_cor = list()
for i in deg_down:
try:
spearman = stats.spearmanr(np.delete(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0], np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))), np.delete(deg_tn_protein.loc[i].values, np.argwhere(np.isnan(pro_met.loc[pro_met_info[pro_met_info['GeneID'] == i].index].iloc[:,84:].values[0]))))
if spearman[1] < 0.05:
if spearman[0] < 0: # p value만 가지고 해야 함 나중에는
deg_down_sig_cor.append(i)
print(i + '\t' + str(spearman[0]) + '\t' + str(spearman[1]))
except (IndexError, ValueError):
pass
# Promoter methylation boxplot
gene = 'SOX2'
df = pd.concat([pro_met.loc[pro_met_info[pro_met_info['GeneID'] == gene].index].iloc[:,84:].T, pd.DataFrame(dmr_t.obs['DMR_leiden2'])], axis=1)
sns.boxplot(data=df, x='DMR_leiden2', y=df.columns[0], palette=color_dict2)
sns.swarmplot(data=df, x='DMR_leiden2', y=df.columns[0], color=".2").set(ylabel=df.columns[0].split('/')[1] + ' promoter methylation')
# Gene boxplot
# (i) 이렇게 하거나
gene = 'TWIST1'
sns.boxplot(data=pd.concat([pd.DataFrame(deg_tn_protein.loc[gene]), pd.DataFrame(dmr_t.obs['DMR_leiden'])], axis=1), x='DMR_leiden', y=gene, palette=color_dict)
sns.swarmplot(data=pd.concat([pd.DataFrame(deg_tn_protein.loc[gene]), pd.DataFrame(dmr_t.obs['DMR_leiden'])], axis=1), x='DMR_leiden', y=gene, color=".2").set(ylabel=gene + ' expression')
# (ii) or 이렇게 해
sns.boxplot(data=pd.concat([pd.DataFrame(rna.iloc[:,84:].loc[gene]), pd.DataFrame(dmr_t.obs['DMR_leiden2'])], axis=1), x='DMR_leiden2', y=gene, palette=color_dict2)
sns.swarmplot(data=pd.concat([pd.DataFrame(rna.iloc[:,84:].loc[gene]), pd.DataFrame(dmr_t.obs['DMR_leiden2'])], axis=1), x='DMR_leiden2', y=gene, color=".2").set(ylabel=gene + ' expression')
# 굳이 둘 간의 mann-whitney U test가 하고 싶다면
df = pd.concat([pd.DataFrame(rna.iloc[:,84:].loc['ESRRG']), pd.DataFrame(dmr_t.obs['DMR_leiden'])], axis=1)
print(stats.mannwhitneyu(df[df['DMR_leiden'] == 'leiden_D']['ESRRG'], df[df['DMR_leiden'] == 'leiden_A']['ESRRG'])[1])
# Leiden D vs A 에서 up 되고 down 된 gene 을 row로 두고 leiden_A, D 에 속하는 샘플들만 hierarchical clustering
fuck1 = deg_tn_protein.loc[:, (deg_tn_protein.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_A'].index)) | (deg_tn_protein.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_D'].index)) ].loc[deg_tn_protein.index.isin(deg_up)]
fuck2 = deg_tn_protein.loc[:, (deg_tn_protein.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_A'].index)) | (deg_tn_protein.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_D'].index)) ].loc[deg_tn_protein.index.isin(deg_down)]
sns.clustermap(fuck1,
method="ward",
metric="euclidean",
z_score=None,
standard_scale=0,
row_cluster=True,
col_cluster=True,
xticklabels=True,
yticklabels=False,
col_colors=list(sample_cluster_color.loc[fuck2.columns]['Colors'].values),
cmap=cmap
)
sns.clustermap(fuck2,
method="ward",
metric="euclidean",
z_score=None,
standard_scale=0,
row_cluster=True,
col_cluster=True,
xticklabels=True,
yticklabels=False,
col_colors=list(sample_cluster_color.loc[fuck2.columns]['Colors'].values),
cmap=cmap
)
# Leiden D vs A 에서 up 되고 down 된 gene의 Promoter 을 row로 두고 leiden_A, D 에 속하는 샘플들만 hierarchical clustering
df = pro_met.loc[pro_met_info[pro_met_info['GeneID'].isin(deg_up)].index].iloc[:,84:]
df = df.loc[:, (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_A'].index)) | (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_D'].index))]
df.index = list(map(lambda x: x.split('/')[1], list(df.index)))
""""""
import scipy.spatial as sp, scipy.cluster.hierarchy as hc
col_dism = 1 - df.corr() # samples
row_dism = 1 - df.T.corr() # promoters
row_linkage = hc.linkage(sp.distance.squareform(row_dism), method='complete')
col_linkage = hc.linkage(sp.distance.squareform(col_dism), method='complete')
sns.clustermap(df,
row_linkage=row_linkage,
col_linkage=col_linkage,
row_cluster=True,
col_cluster=True,
z_score=None,
standard_scale=None,
xticklabels=True,
yticklabels=True,
col_colors=list(sample_cluster_color.loc[df.columns]['Colors'].values),
cmap=cmap
)
""""""
sns.clustermap(df.fillna(0.000000001),
metric='euclidean',
method='ward',
row_cluster=True,
col_cluster=True,
z_score=None,
standard_scale=None,
xticklabels=True,
yticklabels=False,
col_colors=list(sample_cluster_color.loc[df.columns]['Colors'].values),
cmap=cmap
)
df = pro_met.loc[pro_met_info[pro_met_info['GeneID'].isin(deg_down)].index].iloc[:,84:]
df = df.loc[:, (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_A'].index)) | (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_D'].index))]
df.index = list(map(lambda x: x.split('/')[1], list(df.index)))
sns.clustermap(df.fillna(0.000000001),
metric='euclidean',
method='ward',
row_cluster=True,
col_cluster=True,
z_score=None,
standard_scale=None,
xticklabels=True,
yticklabels=False,
col_colors=list(sample_cluster_color.loc[df.columns]['Colors'].values),
cmap=cmap
)
df = pro_met.loc[pro_met_info[pro_met_info['GeneID'].isin(deg_up_sig_cor)].index].iloc[:,84:]
df = df.loc[:, (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_A'].index)) | (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_D'].index))]
df.index = list(map(lambda x: x.split('/')[1], list(df.index)))
sns.clustermap(df.fillna(0.000000001),
metric='euclidean',
method='ward',
row_cluster=True,
col_cluster=True,
z_score=None,
standard_scale=None,
xticklabels=True,
yticklabels=True,
col_colors=list(sample_cluster_color.loc[df.columns]['Colors'].values),
cmap=cmap
)
df = pro_met.loc[pro_met_info[pro_met_info['GeneID'].isin(deg_down_sig_cor)].index].iloc[:,84:]
df = df.loc[:, (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_A'].index)) | (df.columns.isin(dmr_t.obs['DMR_leiden'][dmr_t.obs['DMR_leiden'] == 'leiden_D'].index))]
df.index = list(map(lambda x: x.split('/')[1], list(df.index)))
sns.clustermap(df.fillna(0.000000001),
metric='euclidean',
method='ward',
row_cluster=True,
col_cluster=True,
z_score=None,
standard_scale=None,
xticklabels=True,
yticklabels=True,
col_colors=list(sample_cluster_color.loc[df.columns]['Colors'].values),
cmap=cmap
)
# DMR_leiden clusters and WHO classification
ax = pd.crosstab(dmr_t.obs['WHO'], dmr_t.obs['leiden_r05'], normalize=1).T.plot.bar(stacked=True, color=dict(zip(dmr_t.obs['WHO'].unique(), sns.color_palette("Set2", len(dmr_t.obs['WHO'].unique())))))
ax.legend(loc='upper left', bbox_to_anchor=(1.05, 1.0))
plt.tight_layout()
# DMR_leiden clusters and Lauren's classification
ax = pd.crosstab(df['Lauren'], df['leiden_r05'], normalize=1).T.plot.bar(stacked=True, color=dict(zip(dmr_t.obs['Lauren'].unique(), sns.color_palette("Accent", len(dmr_t.obs['Lauren'].unique())))))
ax.legend(loc='upper left', bbox_to_anchor=(1.05, 1.0))
plt.tight_layout()