-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathDryNB_Survival.R
407 lines (301 loc) · 13.3 KB
/
DryNB_Survival.R
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
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
source("http://bioc.ism.ac.jp/biocLite.R")
biocLite("AnnotationDbi")
install.packages("AnnotationDbi")
install.packages("bit")
biocLite("org.Hs.eg.db")
library(bit)
library(AnnotationDbi)
library(org.Hs.eg.db)
###### Survival package
install.packages("lattice")
library(lattice)
install.packages("survival")
library(survival)
######## plot KM curves
install.packages("reshape2")
library(reshape2)
install.packages("data.table")
library(data.table)
install.packages("zoo")
library("zoo")
install.packages("survminer")
library("survminer")
install.packages("survival")
library("survival")
#### ROC package
# install.packages("gplots")
# library(gplots)
# install.packages("ROCR")
# library(ROCR)
# install.packages("dplyr")
# library("dplyr")
# install.packages("lubridate")
# library(lubridate)
# install.packages("robustbase")
# install.packages("caret")
# library(caret)
install.packages("pROC")
library(pROC)
install.packages("glmnet")
library(glmnet) # manually load this package by downloading and installing from Toll.
setwd("Path/Survival analysis")
#### TCGA TargetedTherapy
Gene_LGG=read.csv("Data/TCGA_LGG_GeneExpression.csv")
rownames(Gene_LGG)=Gene_LGG[,1]
colnames(Gene_LGG)=gsub(".", "-", colnames(Gene_LGG), fixed = TRUE)
Gene_LGG=Gene_LGG[,-1]
Survival_LGG=read.csv("Data/TCGA_LGG_Survival data.csv")
Survival_LGG=Survival_LGG[,c(1,2,4,93)]
rownames(Survival_LGG)=Survival_LGG[,1]
colnames(Survival_LGG)=c('SampleID','OS_event','OS_time','Target_Therapy')
#####
Gene_GBM=read.csv("Data/TCGA_GBM_GeneExpression.csv")
rownames(Gene_GBM)=Gene_GBM[,1]
colnames(Gene_GBM)=gsub(".", "-", colnames(Gene_GBM), fixed = TRUE)
Gene_GBM=Gene_GBM[,-1]
Survival_GBM=read.csv("Data/TCGA_GBM_Survival data.csv")
Survival_GBM=Survival_GBM[,c(1,33,32,109)]
rownames(Survival_GBM)=Survival_GBM[,1]
colnames(Survival_GBM)=c('SampleID','OS_event','OS_time','Target_Therapy')
Gene_TCGA=cbind(Gene_GBM[intersect(rownames(Gene_GBM),rownames(Gene_LGG)),],Gene_LGG[intersect(rownames(Gene_GBM),rownames(Gene_LGG)),])
colnames(Gene_TCGA)=gsub(".", "-", colnames(Gene_TCGA), fixed = TRUE)
Survival_TCGA=rbind(Survival_GBM,Survival_LGG)
Survival_TCGA=Survival_TCGA[Survival_TCGA[,4]=='YES',]
dim(Survival_TCGA)
DNG=read.csv("Data/Quantifing each node.csv") # Differential network genes
Score=as.matrix(DNG[,c(1,9)]) # Importance Score
rownames(Score)=Score[,1]
Score=Score[,-1]
Gene=names(Score)
#####################
#####################
### 100 computations
#####################
#####################
Time=100
sample_ratio=(5:8)*0.1
# AUC_Training_LASSO=matrix(0,length(sample_ratio),Time)
# AUC_Test_LASSO=matrix(0,length(sample_ratio),Time)
# # perf_Training_LASSO=data.frame(Time)
# # perf_Test_LASSO=data.frame(Time)
AUC_Training_wLASSO=matrix(0,length(sample_ratio),Time)
AUC_Test_wLASSO=matrix(0,length(sample_ratio),Time)
sr_ind=0 # index of sample ratio
for (p in sample_ratio)
{
sr_ind=sr_ind+1
for (i in 1:Time)
{
######### Training set - TCGA dataset 1
ind <- sample(2, nrow(Survival_TCGA), replace = TRUE, prob = c(p, 1-p))
Survival_Gene=Gene_TCGA[names(Score),which(ind==1)]
# Survival_Gene_w=Gene_TCGA[,which(ind==1)]
colnames(Survival_Gene)=gsub(".", "-", colnames(Survival_Gene), fixed = TRUE)
Survival=Survival_TCGA[intersect(rownames(Survival_TCGA),colnames(Survival_Gene)),]
Survival=as.matrix(Survival)
Survival_Gene=Survival_Gene[,intersect(rownames(Survival),colnames(Survival_Gene))]
status=as.numeric(Survival[,2])
time=as.numeric(Survival[,3])
Genes=rownames(na.omit(Survival_Gene))
Gene_marker=matrix(as.double(as.matrix(Survival_Gene[Genes,])),dim(Survival_Gene))
time=time[!is.na(status)]
Gene_marker=Gene_marker[,!is.na(status)]
status=status[!is.na(status)]
status[status==1&time>365*3]=0 # 3-year survival status
S=as.numeric(Score)
Gene_marker_w=Gene_marker#*(S-min(S))/max(S) # Importance score-weighted expression
##Dynamic Network-based LASSO
S=as.numeric(Score)
Gene_marker_w=Gene_marker # Importance score-weighted expression
cvfit_wLASSO<-cv.glmnet(t(Gene_marker_w),as.double(status),family="binomial",alpha=1,penalty.factor=S/max(S))
Prediction = predict(cvfit_wLASSO,newx = t(Gene_marker_w),s="lambda.min",type="response")
pred <- prediction(Prediction,status)
# perf_Training_wLASSO[i] <- performance(pred,"tpr","fpr") # calculate probabilities for TPR/FPR for predictions
roc0=roc(status,as.numeric(Prediction))
AUC_Training_wLASSO[sr_ind,i]=as.numeric(auc(status,as.numeric(Prediction)))
######### Validation set - TCGA dataset 2
Survival_Gene=Gene_TCGA[names(Score),]
Survival_Gene=Survival_Gene[,which(ind==2)]
colnames(Survival_Gene)=gsub(".", "-", colnames(Survival_Gene), fixed = TRUE)
Survival=Survival_TCGA[intersect(rownames(Survival_TCGA),colnames(Survival_Gene)),]
Survival=as.matrix(Survival)
Survival_Gene=Survival_Gene[,intersect(rownames(Survival),colnames(Survival_Gene))]
status=as.numeric(Survival[,2])
time=as.numeric(Survival[,3])
Genes=rownames(na.omit(Survival_Gene))
Gene_marker=matrix(as.double(as.matrix(Survival_Gene[Genes,])),dim(Survival_Gene))
time=time[!is.na(status)]
Gene_marker=Gene_marker[,!is.na(status)]
status=status[!is.na(status)]
status[status==1&time>365*3]=0 # 3-year survival status
### Validation for Dynamic Network-based LASSO
S=as.numeric(Score)
Gene_marker_w=Gene_marker#*(S-min(S))/max(S) # Importance score-weighted expression
Prediction = predict(cvfit_wLASSO,newx = t(Gene_marker_w),s="lambda.min",type="response")
pred <- prediction(Prediction,status)
# perf_Test_wLASSO[i] <- performance(pred,"tpr","fpr") # calculate probabilities for TPR/FPR for predictions
roc1=roc(status,as.numeric(Prediction))
AUC_Test_wLASSO[sr_ind,i]=as.numeric(auc(status,as.numeric(Prediction)))
}
}
# AUC_Training_wLASSO
AUC_Test_wLASSO
mean(AUC_Test_wLASSO)
#### Barplots of AUCs_Test dataset of Dynamic Network-based LASSO
dev.new()
boxplot(t(AUC_Training_wLASSO),ylim=c(0,1))
dev.new()
boxplot(t(AUC_Test_wLASSO),ylim=c(0,1))
############### Differential expression profiles of 7 genes in the Normal and Tumor tissues
setwd("Path/Survival analysis")
Gene_LGG=read.csv("Data/TCGA_LGG_GeneExpression.csv")
rownames(Gene_LGG)=Gene_LGG[,1]
colnames(Gene_LGG)=gsub(".", "-", colnames(Gene_LGG), fixed = TRUE)
Gene_LGG=Gene_LGG[,-1]
#
Survival_LGG=read.csv("Data/TCGA_LGG_Survival data.csv")
Survival_LGG=Survival_LGG[,c(1,2,4,87)]
rownames(Survival_LGG)=Survival_LGG[,1]
colnames(Survival_LGG)=c('SampleID','OS_event','OS_time','Target_Therapy')
#####
Gene_GBM=read.csv("Data/TCGA_GBM_GeneExpression.csv")
rownames(Gene_GBM)=Gene_GBM[,1]
colnames(Gene_GBM)=gsub(".", "-", colnames(Gene_GBM), fixed = TRUE)
Gene_GBM=Gene_GBM[,-1]
Survival_GBM=read.csv("Data/TCGA_GBM_Survival data.csv")
Survival_GBM=Survival_GBM[,c(1,33,32,106)]
rownames(Survival_GBM)=Survival_GBM[,1]
colnames(Survival_GBM)=c('SampleID','OS_event','OS_time','Sample Type')
Gene_TCGA=Gene_GBM
colnames(Gene_TCGA)=gsub(".", "-", colnames(Gene_TCGA), fixed = TRUE)
Survival_TCGA=Survival_GBM[intersect(colnames(Gene_GBM),rownames(Survival_GBM)),]
Gene_Normal=Gene_TCGA[,Survival_TCGA[,4]=='Solid Tissue Normal']
Gene_Tumor=Gene_TCGA[,Survival_TCGA[,4]!='Solid Tissue Normal']
dim(Gene_Normal)
dim(Gene_Tumor)
Gene_TCGA=cbind(Gene_LGG,Gene_GBM)
Gene_Tumor=Gene_TCGA[,setdiff(colnames(Gene_TCGA),colnames(Gene_Normal))]
MarkerGenes=c("KIF2C", "CCNA2", "NDC80", "KIF11", "KIF23", "ANLN", "CENPM")
# ID=ensemble2symbol[intersect(rownames(ensemble2symbol),GeneName),]
Marker_Normal=matrix(as.numeric(as.matrix(Gene_Normal[MarkerGenes,])),dim(Gene_Normal[MarkerGenes,]))
Marker_Tumor= matrix(as.numeric(as.matrix(Gene_Tumor[MarkerGenes,])),dim(Gene_Tumor[MarkerGenes,]))
pvalue=matrix(0,1,length(MarkerGenes))
for (i in 1:length(MarkerGenes))
{
Wtest=wilcox.test(Marker_Normal[i,],Marker_Tumor[i,])
pvalue[i]=Wtest$p.value
}
Data=as.data.frame(cbind(Marker_Normal,Marker_Tumor))
Data=t(na.omit(Data))
type=as.factor(c(rep('Normal',5),rep('Tumor',697)))
x=cbind(Data,(type))
# install.packages("vioplot")
library(vioplot)
D=x[x[,8]==1,1:7]
E=x[x[,8]==2,1:7]
dev.new()
vioplot(D[,1],D[,2],D[,3],D[,4],D[,5],D[,6],D[,7],at=seq(1,19,3), names=paste(as.character(MarkerGenes), 'Normal',sep='_'), xlim=c(-0.5,16),col="DarkKhaki", border="gray51", lty=1, lwd=1, rectCol="gray",
colMed="white", pchMed=19)
vioplot(E[,1],E[,2],E[,3],E[,4],E[,5],E[,6],E[,7],at=seq(2,20,3), add=T, names=paste(as.character(MarkerGenes), 'Tumor',sep='_'), xlim=c(-0.5,16),col="DeepSkyBlue3", border="gray51", lty=1, lwd=1, rectCol="gray",
colMed="white", pchMed=19)
##################################################################################################################
##################################################################################################################
################# K-M analysis
################## ensemble ID to gene symbol #################################
ensembl2gene <- toTable(org.Hs.egENSEMBL2EG)
gene2symbol <- toTable(org.Hs.egSYMBOL)
ensemble2symbol <- merge(ensembl2gene, gene2symbol, by = 'gene_id')[2:3]
ensemble2symbol=as.matrix(ensemble2symbol)
rownames(ensemble2symbol)=ensemble2symbol[,2]
###TCGA data
setwd("Path/Survival analysis")
Gene_LGG=read.csv("Data/TCGA_LGG_GeneExpression.csv")
rownames(Gene_LGG)=Gene_LGG[,1]
colnames(Gene_LGG)=gsub(".", "-", colnames(Gene_LGG), fixed = TRUE)
Gene_LGG=Gene_LGG[,-1]
Survival_LGG=read.csv("Data/TCGA_LGG_Survival data.csv")
Survival_LGG=Survival_LGG[,c(1,2,4)]
rownames(Survival_LGG)=Survival_LGG[,1]
colnames(Survival_LGG)=c('SampleID','OS_event','OS_time')
#####
Gene_GBM=read.csv("Data/TCGA_GBM_GeneExpression.csv")
rownames(Gene_GBM)=Gene_GBM[,1]
colnames(Gene_GBM)=gsub(".", "-", colnames(Gene_GBM), fixed = TRUE)
Gene_GBM=Gene_GBM[,-1]
Survival_GBM=read.csv("Data/TCGA_GBM_Survival data.csv")
Survival_GBM=Survival_GBM[,c(1,33,32)]
rownames(Survival_GBM)=Survival_GBM[,1]
colnames(Survival_GBM)=c('SampleID','OS_event','OS_time')
Gene_TCGA=cbind(Gene_GBM[intersect(rownames(Gene_GBM),rownames(Gene_LGG)),],Gene_LGG[intersect(rownames(Gene_GBM),rownames(Gene_LGG)),])
colnames(Gene_TCGA)=gsub(".", "-", colnames(Gene_TCGA), fixed = TRUE)
Survival_TCGA=rbind(Survival_GBM,Survival_LGG)
######################## Select genes for signature #####################################
MNB=c("KIF2C", "CCNA2", "NDC80", "KIF11", "KIF23", "ANLN", "CENPM") # Cell Cycle-related gene signature
Survival_Gene=Gene_TCGA[t(MNB),]
colnames(Survival_Gene)=gsub(".", "-", colnames(Survival_Gene), fixed = TRUE)
Survival=Survival_TCGA[intersect(rownames(Survival_TCGA),colnames(Survival_Gene)),]
Survival=as.matrix(Survival)
Survival_Gene=Survival_Gene[,intersect(rownames(Survival),colnames(Survival_Gene))]
### Training dataset
status=as.numeric(Survival[,2])
time=as.numeric(Survival[,3])
Genes=rownames(na.omit(Survival_Gene))
Gene_marker=matrix(as.double(as.matrix(Survival_Gene[Genes,])),dim(Survival_Gene))
time=time[!is.na(status)]
Gene_marker=Gene_marker[,!is.na(status)]
status=status[!is.na(status)]
### COX Model
surv=Surv(as.double(time),as.double(status))
fit <- coxph(surv ~ t(Gene_marker), method="efron")
A=coef(fit)
#
# surv=Surv(as.double(time),as.double(status))
#
# cvfit<-cv.glmnet(t(Gene_marker),surv,family="cox",alpha=1)
# plot(cvfit)
# A=coef(cvfit, s = "lambda.min")
# A=coef(cvfit, s="lambda.1se")
# which(A!=0)
# sum(A!=0)
# MNB[which(A!=0)]
#
# fit <- coxph(Surv(time, status) ~ t(Gene_marker[which(A!=0),]), method="breslow")
######## ROC for Training set
predicted=matrix(0,1,dim(Gene_marker)[2])
for (i in 1:dim(Gene_marker)[2])
{
predicted[i]=sum(A*Gene_marker[,i])
}
predicted0=as.numeric(predicted)
predicted0=predicted0[!is.na(status)]
status0=as.matrix(na.omit(status))
pred <- prediction(predicted0,status0)
perf <- performance(pred,"tpr","fpr") # calculate probabilities for TPR/FPR for predictions
performance(pred,"auc") # shows calculated AUC for model
dev.new()
plot(perf,colorize=FALSE, col="DarkCyan") # plot ROC curve
lines(c(0,1),c(0,1),col = "gray", lty = 4 )
roc0=roc((status0),predicted0,smooth=TRUE,plot=TRUE)
roc0
AUC=auc(roc0,predicted0)
AUC
## optimal combination
opt <- which.max(rowSums(cbind(roc0$sensitivities,roc0$specificities)))
## optimal cut-off point
sort(predicted0,F)[opt]
########## K-M survival curves for MNB in TCGA datasets
groups=matrix(0,1,length(status))
groups[predicted<=sort(predicted,F)[opt]]=1
groups[predicted>sort(predicted,F)[opt]]=2
# groups[predicted<=median(predicted)]=1
# groups[predicted>median(predicted)]=2
groups=t(groups)
groups=as.numeric(groups)
fit<- survfit(Surv(time/365, status) ~ groups, data = as.data.frame(Survival_Gene))
# Drawing survival curves
dev.new()
ggsurvplot(fit, legend = c(0.2, 0.2))
# Add risk table
# and change risk table y text colors by strata
ggsurvplot(fit, pval = TRUE, conf.int = TRUE,
risk.table = TRUE, risk.table.y.text.col = TRUE)