Survival Analysis with Genes

Calculation of Three Genes Expression using Excel

• 3 genes expression average G2 = AVERAGE(D2, E2, F2)

• Zscore H2 = (G2-AVERAGE($G\(2:\)G\(100))/STDEV(\)G\(2:\)G$100)

• zrank 25 I2 = IF(RANK(H2,$H\(2:\)H$100,1) > 99*0.75,“High”,“Low”)

• zrank 25-75 J2 = IF(RANK(H2,$H\(2:\)H$100,1) > 990.75,“High”, IF(RANK(H2, $H\(2:\)H$100, 1) <990.25, “Low”,“”))

• zrank median K2 = IF(H2>MEDIAN($H\(2:\)H$100),“High”, “Low”)

library(survival)
library(survminer)

## Loading required package: ggplot2

## Loading required package: ggpubr

## Loading required package: magrittr

df = read.csv("OS_NEK7.csv", header = TRUE, sep = ",")
head(df)

##            sample RFS RFS.time OS OS.time zrank.median.3genes NEK7.rank
## 1 TCGA-EJ-5516-01   0     1889  0    1889                High      High
## 2 TCGA-FC-7961-01   0      469  0     469                High      High
## 3 TCGA-G9-6363-01   0     1378  0    1378                High      High
## 4 TCGA-EJ-5517-01   0     1889  0    1889                High      High
## 5 TCGA-G9-6362-01  NA       NA  0    1443                High      High
## 6 TCGA-EJ-5526-01   1     1909  0    1909                High      High
##   NR3C1.rank UNC45A.rank
## 1       High         Low
## 2       High         Low
## 3       High         Low
## 4       High         Low
## 5       High         Low
## 6       High         Low

fit.OS.NEK7 = survfit(Surv(OS.time, OS) ~ NEK7.rank, data = df)
fit.RFS.NEK7 = survfit(Surv(RFS.time, RFS) ~ NEK7.rank, data = df)

ggsurvplot(fit.OS.NEK7, pval=TRUE, legend="right", legend.labs = c("above median", "below median"), xlab = "Time in days", title = "OS - NEK7")

ggsurvplot(fit.RFS.NEK7, pval=TRUE, legend="right", legend.labs = c("above median", "below median"), xlab = "Time in days", title = "RFS - NEK7")

fit.OS.3genes = survfit(Surv(OS.time, OS) ~ zrank.median.3genes, data = df)
fit.RFS.3genes = survfit(Surv(RFS.time, RFS) ~ zrank.median.3genes, data = df)

ggsurvplot(fit.RFS.3genes, pval=TRUE, legend="right", legend.labs = c("above median", "below median"), xlab = "Time in days", title = "RFS - NEK7, NR3C1, UNC45A")

ggsurvplot(fit.OS.3genes, pval=TRUE, legend="right", legend.labs = c("above median", "below median"), xlab = "Time in days", title = "OS - NEK7, NR3C1, UNC45A")