-
Notifications
You must be signed in to change notification settings - Fork 1
/
val_neoadj_sign.Rmd
211 lines (168 loc) · 13.9 KB
/
val_neoadj_sign.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
---
title: "Validation breast cancer metagene score"
author: "M. Hoogstraat"
date: "3/24/2022"
output: html_document
---
# R code to validate the metagene score from M. Hoogstraat et al., in an independent dataset
The new dataset should contain normalized gene expression data in log scale.
These values can then be transformed using the (modified) Z-score, for example:
```{r z_transform_example, eval=F}
exprs_z <- 0.6745 * (exprs - apply(exprs, 1, median)) / abs(apply(exprs, 1, mad))
```
...where 'exprs' is a dataframe containing ONLY the normalized, log scaled expression data, with samples in columns and genes in rows.
Column headers should contain sample identifiers.
Rownames should contain gene identifiers (HUGO symbol, ensembl gene id, entrez id).
As an example, the R source code to validate the signature in the TCGA breast cancer dataset is provided below.
# Validation of the metagene score in the [TCGA BRCA 2018 dataset](https://www.sciencedirect.com/science/article/pii/S0092867418303027)
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```
#### Loading required libraries
```{r loading_libs, message=F, warning=F}
library(survival)
```
#### Loading required files
```{r loading_data}
rna_zscores <- read.delim("data_RNA_Seq_v2_mRNA_median_Zscores.txt") #Source data not provided
clindat <- read.delim("data_clinical_patient.txt",comment.char="#") #Source data not provided
clindat$colnames <- paste(gsub("-",".",clindat$PATIENT_ID),".01",sep="")
metagene_genes<-read.delim("metagene_genelist.txt")
rownames(metagene_genes) <- metagene_genes$gene_name
```
#### Testing the amount of genes from original signature present in new data
```{r test_overlap}
knitr::kable(table(metagene_genes$gene_name[which(metagene_genes$process %in% c('proliferation','immune response','ECM associated'))] %in% rna_zscores$Hugo_Symbol), caption="Genes from signature present in new data")
```
## BRCA_Basal
#### Obtain gene expression levels from signature genes
```{r get_genes_Basal}
rna_zscores_basal <- as.data.frame(rbind(
rna_zscores[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "proliferation")]),
as.vector(clindat$colnames[which(clindat$SUBTYPE == "BRCA_Basal")])],
rna_zscores[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "immune response")]),
as.vector(clindat$colnames[which(clindat$SUBTYPE == "BRCA_Basal")])]
))
rna_zscores_basal <- as.data.frame(rbind(
rna_zscores_basal,
rna_zscores[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "ECM associated")]),
as.vector(clindat$colnames[which(clindat$SUBTYPE == "BRCA_Basal")])]
))
rownames(rna_zscores_basal) <- c(
as.vector(rna_zscores$Hugo_Symbol[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "proliferation")])]),
as.vector(rna_zscores$Hugo_Symbol[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "immune response")])]),
as.vector(rna_zscores$Hugo_Symbol[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "ECM associated")])])
)
```
#### Calculate metagenes and cutoffs
```{r get_metagenes_Basal}
clindat_basal <- clindat[which(clindat$SUBTYPE == "BRCA_Basal"),]
clindat_basal$Proliferation <- apply(rna_zscores_basal[as.vector(metagene_genes$gene_name[which(metagene_genes$process == "proliferation" & metagene_genes$gene_name %in% rownames(rna_zscores_basal))]),],2,median)
clindat_basal$Immune <- apply(rna_zscores_basal[as.vector(metagene_genes$gene_name[which(metagene_genes$process == "immune response" & metagene_genes$gene_name %in% rownames(rna_zscores_basal))]),],2,median)
clindat_basal$ECM <- apply(rna_zscores_basal[as.vector(metagene_genes$gene_name[which(metagene_genes$process == "ECM associated" & metagene_genes$gene_name %in% rownames(rna_zscores_basal))]),],2,median)
clindat_basal$cutoffs <- "other"
clindat_basal$cutoffs[which(clindat_basal$Proliferation >= median(clindat_basal$Proliferation) & clindat_basal$Immune >= median(clindat_basal$Immune) & clindat_basal$ECM < median(clindat_basal$ECM))] <- "HHL"
clindat_basal$cutoffs[which(clindat_basal$Proliferation < median(clindat_basal$Proliferation) & clindat_basal$Immune < median(clindat_basal$Immune) & clindat_basal$ECM >= median(clindat_basal$ECM))] <- "LLH"
```
#### Process clinical data
```{r parse_clindat_Basal}
clindat_basal$PFS_stat_bin <- 0
clindat_basal$PFS_stat_bin[which(clindat_basal$PFS_STATUS == "PROGRESSION")] <- 1
clindat_basal$n_stage <- 1
clindat_basal$n_stage[grep("N0",clindat_basal$PATH_N_STAGE)] <- 0
clindat_basal$t_stage_bin <- NA
clindat_basal$t_stage_bin[grep("T[12]",clindat_basal$PATH_T_STAGE)] <- "1_2"
clindat_basal$t_stage_bin[grep("T[34]",clindat_basal$PATH_T_STAGE)] <- "3_4"
```
#### Calculate association metagene score vs progression-free survival in various subgroups
```{r associate_clindat_metagene_Basal}
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal)
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$n_stage == 1 | clindat_basal$t_stage_bin == "3_4"),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$n_stage == 0 ),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$n_stage == 1 ),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$t_stage_bin == "1_2"),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$t_stage_bin == "3_4"),])
```
#### Visualize association metagene score vs progression-free survival in various subgroups
```{r visualize_assoc_Basal, fig.height=10}
par(mfrow=c(3,2))
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal), mark.time=T,col=c("orange","blue","grey"),main="All_basal",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,20))
legend('bottomleft',lwd=2,col=c('blue','orange','grey'), legend=c('HHL','LLH','other'),cex=1.5)
text(15,0.2,labels="P = 0.2",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$n_stage == 1 | clindat_basal$t_stage_bin == "3_4"),]), mark.time=T,col=c("orange","blue","grey"),main="Basal, N+ or T-stage 3 or 4",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,20))
text(15,0.2,labels="P = 0.02",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$n_stage == 0 ),]), mark.time=T,col=c("orange","blue","grey"),main="Basal, N-",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,20))
text(15,0.2,labels="P = 0.7",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$n_stage == 1 ),]), mark.time=T,col=c("orange","blue","grey"),main="Basal, N+",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,20))
text(15,0.2,labels="P = 0.004",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$t_stage_bin == "1_2" ),]), mark.time=T,col=c("orange","blue","grey"),main="Basal, T-stage 1 or 2",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,20))
text(15,0.2,labels="P = 0.8",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_basal[which(clindat_basal$t_stage_bin == "3_4" ),]), mark.time=T,col=c("orange","blue","grey"),main="Basal, T-stage 3 or 4",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,20))
text(15,0.2,labels="P = 0.3",cex=2)
```
## BRCA_Luminal
#### Obtain gene expression levels from signature genes
```{r get_genes_Lum}
rna_zscores_lum <- as.data.frame(rbind(
rna_zscores[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "proliferation")]),
as.vector(clindat$colnames[which(clindat$SUBTYPE %in% c("BRCA_LumA", "BRCA_LumB"))])],
rna_zscores[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "immune response")]),
as.vector(clindat$colnames[which(clindat$SUBTYPE %in% c("BRCA_LumA", "BRCA_LumB"))])]
))
rna_zscores_lum <- as.data.frame(rbind(
rna_zscores_lum,
rna_zscores[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "ECM associated")]),
as.vector(clindat$colnames[which(clindat$SUBTYPE %in% c("BRCA_LumA", "BRCA_LumB"))])]
))
rownames(rna_zscores_lum) <- c(
as.vector(rna_zscores$Hugo_Symbol[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "proliferation")])]),
as.vector(rna_zscores$Hugo_Symbol[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "immune response")])]),
as.vector(rna_zscores$Hugo_Symbol[which(rna_zscores$Hugo_Symbol %in% metagene_genes$gene_name[which(metagene_genes$process == "ECM associated")])])
)
```
#### Calculate metagenes and cutoffs
```{r get_metagenes_Lum}
clindat_lum <- clindat[which(clindat$SUBTYPE %in% c("BRCA_LumA", "BRCA_LumB")),]
clindat_lum$PFS_stat_bin <- 0
clindat_lum$PFS_stat_bin[which(clindat_lum$PFS_STATUS == "PROGRESSION")] <- 1
clindat_lum$Proliferation <- apply(rna_zscores_lum[as.vector(metagene_genes$gene_name[which(metagene_genes$process == "proliferation" & metagene_genes$gene_name %in% rownames(rna_zscores_lum))]),],2,median)
clindat_lum$Immune <- apply(rna_zscores_lum[as.vector(metagene_genes$gene_name[which(metagene_genes$process == "immune response" & metagene_genes$gene_name %in% rownames(rna_zscores_lum))]),],2,median)
clindat_lum$ECM <- apply(rna_zscores_lum[as.vector(metagene_genes$gene_name[which(metagene_genes$process == "ECM associated" & metagene_genes$gene_name %in% rownames(rna_zscores_lum))]),],2,median)
```
#### Process clinical data
```{r parse_clindat_Lum}
clindat_lum$cutoffs <- "other"
clindat_lum$cutoffs[which(clindat_lum$Proliferation >= median(clindat_lum$Proliferation) & clindat_lum$Immune >= median(clindat_lum$Immune) & clindat_lum$ECM < median(clindat_lum$ECM))] <- "HHL"
clindat_lum$cutoffs[which(clindat_lum$Proliferation < median(clindat_lum$Proliferation) & clindat_lum$Immune < median(clindat_lum$Immune) & clindat_lum$ECM >= median(clindat_lum$ECM))] <- "LLH"
clindat_lum$n_stage <- 1
clindat_lum$n_stage[grep("N0",clindat_lum$PATH_N_STAGE)] <- 0
clindat_lum$t_stage_bin <- NA
clindat_lum$t_stage_bin[grep("T[12]",clindat_lum$PATH_T_STAGE)] <- "1_2"
clindat_lum$t_stage_bin[grep("T[34]",clindat_lum$PATH_T_STAGE)] <- "3_4"
```
#### Calculate association metagene score vs progression-free survival in various subgroups
```{r associate_clindat_metagene_Lum}
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum)
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$n_stage == 1 | clindat_lum$t_stage_bin == "3_4"),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$n_stage == 0 ),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$n_stage == 1 ),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$t_stage_bin == "1_2" ),])
survdiff(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$t_stage_bin == "3_4" ),])
```
#### Visualize association metagene score vs progression-free survival in various subgroups
```{r visualize_assoc_Lum, fig.height=10}
par(mfrow=c(3,2))
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum), mark.time=T,col=c("orange","blue","grey"),main="All_lum",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,14))
legend('bottomleft',lwd=2,col=c('blue','orange','grey'), legend=c('HHL','LLH','other'),cex=1.5)
text(10,0.2,labels="P = 0.09",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$n_stage == 1 | clindat_lum$t_stage_bin == "3_4"),]), mark.time=T,col=c("orange","blue","grey"),main="Luminal, N+ or T-stage 3 or 4",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,14))
text(10,0.2,labels="P = 0.1",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$n_stage == 0 ),]), mark.time=T,col=c("orange","blue","grey"),main="Luminal, N-",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,14))
text(10,0.2,labels="P = 0.8",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$n_stage == 1 ),]), mark.time=T,col=c("orange","blue","grey"),main="Luminal, N+",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,14))
text(10,0.2,labels="P = 0.1",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$t_stage_bin == "1_2" ),]), mark.time=T,col=c("orange","blue","grey"),main="Luminal, T-stage 1 or 2",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,14))
text(10,0.2,labels="P = 0.5",cex=2)
plot(survfit(Surv(PFS_MONTHS/12, PFS_stat_bin) ~ cutoffs, data=clindat_lum[which(clindat_lum$t_stage_bin == "3_4" ),]), mark.time=T,col=c("orange","blue","grey"),main="Luminal, T-stage 3 or 4",xlab="Progression-free survival (years)",ylab="Proportion of patients at risk",lwd=2, cex.lab=1.5, cex.axis=1.5,cex.main=1.5,xlim=c(0,14))
text(10,0.2,labels="P = 0.1",cex=2)
```