forked from xuezhangzhi/zzov
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathFig4_Cox.Rmd
322 lines (239 loc) · 14.5 KB
/
Fig4_Cox.Rmd
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
---
title: "COX_Fig4"
author: "Yan Zhou"
date: "2024-04-01"
output: html_document
---
```{r setup, include=FALSE}
rm(list = ls())
pacman::p_load(readxl, openxlsx, #Load One or More Packages
reshape2, magrittr, tidyverse, Hmisc,
ggrepel,
ggpubr, ggsignif, ggridges,
RColorBrewer, randomcoloR,
gridExtra, ggplotify,
ggupset, VennDiagram,
FactoMineR, factoextra, Rtsne, umap,
pheatmap, ComplexHeatmap, vioplot,
circlize, RCircos, corrplot,
Biobase, survival, survminer, magrittr,
ConsensusClusterPlus)
ge.split <- function(data,split,which=1){
sapply(data,function(v){strsplit(v,split)[[1]][which]})
}
```
```{r import data}
pinfo = read.xlsx("20210421ZZOV_patient_923sample_info.xlsx", sheet = 1)
sinfo = read.xlsx("20210421ZZOV_patient_923sample_info.xlsx", sheet = 2)
pm = read.table("20210615ZZOV_923sample_quanNor_NA=0.8min_diann_prot.txt") # updated matrix
pm_NA = read.table("20210413ZZOV_923sample_withNA_diann_prot.txt")
```
```{r filter patients info}
# primary patients
pinfo_pry = pinfo[pinfo$Group == "Primary carcinoma" &
pinfo$Histology8 == "HS" &
(pinfo$modified_Tumor_stage == "3" |
pinfo$modified_Tumor_stage == "4" ) &
(pinfo$`Frequency.of.chemotherapy(A:-2,B:3-5,C:6-8,D:≥9)` == "C" |
pinfo$`Frequency.of.chemotherapy(A:-2,B:3-5,C:6-8,D:≥9)` == "D") &
(pinfo$Bevacizumab.or.PARP.inhibitor == "No" |
pinfo$Bevacizumab.or.PARP.inhibitor == "NO" |
pinfo$Bevacizumab.or.PARP.inhibitor == "NOT available") &
!is.na(pinfo$`Recurrence.free.survival(RFS)/month`) &
!pinfo$`Recurrence.free.survival(RFS)/month` == "NOT available" ,]
bcr = sinfo$bcr_patient_barcode[match(names(pm), sinfo$Sample_name)]
# only patients that still in the protein matrix
pinfo1_pry = pinfo_pry[pinfo_pry$Bcr_patient_barcode %in% bcr,]
sinfo_pry = sinfo[sinfo$bcr_patient_barcode %in% pinfo1_pry$Bcr_patient_barcode,]
# relapsing patients
pinfo_rlps = pinfo[pinfo$Group == "Relapsing carcinoma" &
pinfo$Histology8 == "HS" &
!is.na(pinfo$`Recurrence.free.survival(RFS)/month`) &
!pinfo$`Recurrence.free.survival(RFS)/month` == "NOT available" &
(pinfo$`Frequency.of.chemotherapy(A:-2,B:3-5,C:6-8,D:≥9)` == "C" |
pinfo$`Frequency.of.chemotherapy(A:-2,B:3-5,C:6-8,D:≥9)` == "D") &
(pinfo$Bevacizumab.or.PARP.inhibitor == "No" |
pinfo$Bevacizumab.or.PARP.inhibitor == "NO" |
pinfo$Bevacizumab.or.PARP.inhibitor == "NOT available" |
is.na(pinfo$Bevacizumab.or.PARP.inhibitor)),]
pinfo1_rlps = pinfo_rlps[pinfo_rlps$Bcr_patient_barcode %in% bcr,]
sinfo_rlps = sinfo[sinfo$bcr_patient_barcode %in% pinfo1_rlps$Bcr_patient_barcode,]
# neo-adjuvant patients
pinfo_neo = pinfo[pinfo$Group == "Carcinoma with neoadjuvant therapy" &
pinfo$Histology8 == "HS" &
(pinfo$modified_Tumor_stage == "3" |
pinfo$modified_Tumor_stage == "4" ) &
!is.na(pinfo$`Recurrence.free.survival(RFS)/month`) &
!pinfo$`Recurrence.free.survival(RFS)/month` == "NOT available" &
(pinfo$`Frequency.of.chemotherapy(A:-2,B:3-5,C:6-8,D:≥9)` == "C" |
pinfo$`Frequency.of.chemotherapy(A:-2,B:3-5,C:6-8,D:≥9)` == "D") &
(pinfo$Bevacizumab.or.PARP.inhibitor == "No" |
pinfo$Bevacizumab.or.PARP.inhibitor == "NO" |
pinfo$Bevacizumab.or.PARP.inhibitor == "NOT available" |
is.na(pinfo$Bevacizumab.or.PARP.inhibitor)),]
pinfo1_neo = pinfo_neo[pinfo_neo$Bcr_patient_barcode %in% bcr,]
sinfo_neo = sinfo[sinfo$bcr_patient_barcode %in% pinfo1_neo$Bcr_patient_barcode,]
```
```{r data preprocessing}
## NA% threshold = 70%
na_rate = rowSums(is.na(pm_NA)) / ncol(pm_NA)
na_thres = 0.70
pm_NA70 <- pm_NA[na_rate < na_thres,]
pm1 = pm[row.names(pm_NA70),]
## Normalization according to age
# INT function
int <- function(x){
#inverse normal transformation
qnorm((rank(x,na.last="keep")-0.5)/sum(!is.na(x)))
}
# data preparation
names(sinfo)[2] = "Bcr_patient_barcode"
pinfo = pinfo[pinfo$Bcr_patient_barcode %in% sinfo$Bcr_patient_barcode,]
sinfo_merged = left_join(sinfo, pinfo[,c("Bcr_patient_barcode","Age_at_diagnosis",
"Doctor","Tumor_residual_disease((no.residual),1(≤1㎝),2(＞1㎝),Not.available)",
"Metastasis.of.lymph",
"Sampling.location(A:uterine.appendages;B:pelvic.cavity;C:abdominal.cavity;D:Distant.metastasis)",
"Number.of.peritoneal.chemotherapy","CA125.before.the.treatment","HE4.before.the.treatment", "CA125.of.drug.resistance.before.last.chemotherapy",
"Recurrence.free.survival(RFS)/month",
"Recurrent.disease")], by = 'Bcr_patient_barcode')
pm2 = t(pm1) %>% data.frame()
pm2 = rownames_to_column(pm2, var = "Sample_name")
pro_merged = left_join(pm2,sinfo_merged, by = "Sample_name" )
## rank-based INT
# Age residuals + standardization
res_stan <- function(column){
lm.fit <- lm(formula = column~Age_at_diagnosis,data=pro_merged)
res <- lm.fit$residuals
res <- int(res)
}
pro_res <- sapply(pro_merged[,2:7720], res_stan)
row.names(pro_res) = pro_merged$Sample_name
pro_res = pro_res[sinfo_merged$Sample_name,] %>% t() %>% data.frame()
```
```{r univariate COX}
##pry
raw_mat_pry <- pro_res[, sinfo_pry$Sample_name]
prot_pry_impMin = raw_mat_pry %>% t() %>% as.data.frame()
sample_pry = cbind(pinfo1_pry$Patient_ID, pinfo1_pry$`Recurrence.free.survival(RFS)/month`, pinfo1_pry$Recurrent.disease) %>% as.data.frame()
names(sample_pry) = c("Patient_ID", "RFS", "recur")
sample_pry$recur = factor(sample_pry$recur, labels = c(0,1)) %>% as.numeric()
nm2 = str_sub(row.names(prot_pry_impMin), end = -4)
prot_pry_impMin$RFS = sample_pry$RFS[match(nm2, sample_pry$Patient_ID)] %>% as.numeric()
prot_pry_impMin$recur = sample_pry$recur[match(nm2, sample_pry$Patient_ID)] %>% as.numeric()
### rlps
raw_mat_rlps <- pro_res[, sinfo_rlps$Sample_name]
prot_rlps_impMin <- raw_mat_rlps %>% t() %>% as.data.frame()
sample_rlps = cbind(pinfo1_rlps$Patient_ID, pinfo1_rlps$`Recurrence.free.survival(RFS)/month`, pinfo_rlps$Recurrent.disease) %>% as.data.frame()
names(sample_rlps) = c("Patient_ID", "RFS", "recur")
sample_rlps$recur = factor(sample_rlps$recur, labels = c(0,1)) %>% as.numeric()
nm2 = str_sub(row.names(prot_rlps_impMin), end = -4)
prot_rlps_impMin$RFS = sample_rlps$RFS[match(nm2, sample_rlps$Patient_ID)] %>% as.numeric()
prot_rlps_impMin$recur = sample_rlps$recur[match(nm2, sample_rlps$Patient_ID)] %>% as.numeric()
# ## neo matrix
raw_mat_neo <- pro_res[, sinfo_neo$Sample_name]
prot_neo_impMin <- raw_mat_neo %>% t() %>% as.data.frame()
sample_neo = cbind(pinfo1_neo$Patient_ID, pinfo1_neo$`Recurrence.free.survival(RFS)/month`, pinfo1_neo$Recurrent.disease) %>% as.data.frame()
names(sample_neo) = c("Patient_ID", "RFS", "recur")
sample_neo$recur = factor(sample_neo$recur, labels = c(0,1)) %>% as.numeric()
nm2 = str_sub(row.names(prot_neo_impMin), end = -4)
prot_neo_impMin$RFS = sample_neo$RFS[match(nm2, sample_neo$Patient_ID)] %>% as.numeric()
prot_neo_impMin$recur = sample_neo$recur[match(nm2, sample_neo$Patient_ID)] %>% as.numeric()
### calculate
##pry
# names(prot_pry_impMin) = gsub("-", "_", names(prot_pry_impMin)) # will report errors if there is a digit after the hyphen-
# names(prot_pry_impMin) = gsub("/", "_", names(prot_pry_impMin))
covariates <- names(prot_pry_impMin[,1:(ncol(prot_pry_impMin)-2)]) # protein names
univ_formulas <- sapply(covariates,
function(x) as.formula(paste('Surv(RFS, recur)~',x)))
univ_models <- lapply(univ_formulas, function(x){coxph(x, data = prot_pry_impMin)})
univ_results <- lapply(univ_models,
function(x){
x <- summary(x)
wald.pvalue <- signif(x$wald['pvalue'], digits = 2) %>% as.numeric()
wald.test <- signif(x$wald["test"], digits = 2)
log.pvalue <- signif(x$logtest['pvalue'], digits = 2) %>% as.numeric()
log.test <- signif(x$logtest["test"], digits = 2)
beta <- signif(x$coef[1], digits=2);
HR <- signif(x$coef[2], digits=2);
HR.confint.lower <- signif(x$conf.int[,"lower .95"],2)
HR.confint.upper <- signif(x$conf.int[,"upper .95"],2)
HR <- paste0(HR, " (",
HR.confint.lower, "-", HR.confint.upper, ")")
res<-c(beta, HR, wald.test, wald.pvalue,log.test, log.pvalue)
names(res)<-c("beta", "HR (95% CI for HR)",
"wald.test", "wald.pvalue", "log.test", "log.pvalue")
return(res)
})
res_pry <- t(as.data.frame(univ_results, check.names = FALSE)) %>% as.data.frame()
res2_pry <- res_pry[!is.na(res_pry$log.pvalue),]
res2_pry$log.pvalue = as.numeric(res2_pry$log.pvalue)
res2_pry$log.p_adjust = p.adjust(res2_pry$log.pvalue, method = "BH")
min(res2_pry$log.p_adjust) # min = 0.007719
write.xlsx(res2_pry, file = "source_data/COX_pry_int.xlsx", rowNames = T)
## rlps
names(prot_rlps_impMin) = gsub("-", "_", names(prot_rlps_impMin)) # will report errors if there is a digit after the hyphen-
names(prot_rlps_impMin) = gsub("/", "_", names(prot_rlps_impMin))
covariates <- names(prot_rlps_impMin[,1:(ncol(prot_rlps_impMin)-2)]) # protein names
univ_formulas <- sapply(covariates,
function(x) as.formula(paste('Surv(RFS, recur)~',x)))
univ_models <- lapply(univ_formulas, function(x){coxph(x, data = prot_rlps_impMin)})
univ_results <- lapply(univ_models,
function(x){
x <- summary(x)
wald.pvalue <- signif(x$wald['pvalue'], digits = 2) %>% as.numeric()
wald.test <- signif(x$wald["test"], digits = 2)
log.pvalue <- signif(x$logtest['pvalue'], digits = 2) %>% as.numeric()
log.test <- signif(x$logtest["test"], digits = 2)
beta <- signif(x$coef[1], digits=2);
HR <- signif(x$coef[2], digits=2);
HR.confint.lower <- signif(x$conf.int[,"lower .95"],2)
HR.confint.upper <- signif(x$conf.int[,"upper .95"],2)
HR <- paste0(HR, " (",
HR.confint.lower, "-", HR.confint.upper, ")")
res<-c(beta, HR, wald.test, wald.pvalue,log.test, log.pvalue)
names(res)<-c("beta", "HR (95% CI for HR)",
"wald.test", "wald.pvalue", "log.test", "log.pvalue")
return(res)
})
res_rlps <- t(as.data.frame(univ_results, check.names = FALSE)) %>% as.data.frame()
res2_rlps <- res_rlps[!is.na(res_rlps$log.pvalue),]
res2_rlps$log.pvalue = as.numeric(res2_rlps$log.pvalue)
res2_rlps$log.p_adjust = p.adjust(res2_rlps$log.pvalue, method = "BH")
min(res2_rlps$log.p_adjust) # 0.9582663
write.xlsx(res2_rlps, file = "source_data/COX_rlps_int.xlsx", rowNames = T)
res3_rlps_0.05 <- res2_rlps[res2_rlps$log.pvalue < 0.05,]
res3_rlps_0.01 <- res2_rlps[res2_rlps$log.pvalue < 0.01,]
res3_rlps_0.001 <- res2_rlps[res2_rlps$log.pvalue < 0.001,]
res3_rlps_0.005 <- res2_rlps[res2_rlps$log.pvalue < 0.005,]
###neo
names(prot_neo_impMin) = gsub("-", "_", names(prot_neo_impMin)) # will report errors if there is a digit after the hyphen-
names(prot_neo_impMin) = gsub("/", "_", names(prot_neo_impMin))
covariates <- names(prot_neo_impMin[,1:(ncol(prot_neo_impMin)-2)]) # protein names
univ_formulas <- sapply(covariates,
function(x) as.formula(paste('Surv(RFS, recur)~',x)))
univ_models <- lapply(univ_formulas, function(x){coxph(x, data = prot_neo_impMin)})
univ_results <- lapply(univ_models,
function(x){
x <- summary(x)
wald.pvalue <- signif(x$wald['pvalue'], digits = 2) %>% as.numeric()
wald.test <- signif(x$wald["test"], digits = 2)
log.pvalue <- signif(x$logtest['pvalue'], digits = 2) %>% as.numeric()
log.test <- signif(x$logtest["test"], digits = 2)
beta <- signif(x$coef[1], digits=2);
HR <- signif(x$coef[2], digits=2);
HR.confint.lower <- signif(x$conf.int[,"lower .95"],2)
HR.confint.upper <- signif(x$conf.int[,"upper .95"],2)
HR <- paste0(HR, " (",
HR.confint.lower, "-", HR.confint.upper, ")")
res<-c(beta, HR, wald.test, wald.pvalue,log.test, log.pvalue)
names(res)<-c("beta", "HR (95% CI for HR)",
"wald.test", "wald.pvalue", "log.test", "log.pvalue")
return(res)
})
res_neo <- t(as.data.frame(univ_results, check.names = FALSE)) %>% as.data.frame()
res2_neo <- res_neo[!is.na(res_neo$log.pvalue),]
res2_neo$log.pvalue = as.numeric(res2_neo$log.pvalue)
res2_neo$log.p_adjust = p.adjust(res2_neo$log.pvalue, method = "BH")
min(res2_neo$log.p_adjust) ## 0.3049005
write.xlsx(res2_neo, file = "source_data/COX_neo_int.xlsx", rowNames = T)
```