泛癌临床相关性分析

#if (!requireNamespace("BiocManager", quietly = TRUE))
#    install.packages("BiocManager")
#BiocManager::install("limma")

#install.packages("ggpubr")

library(limma)
library(ggpubr)
library(dplyr)
setwd("H:\\BaiduNetdiskDownload\\pan-cancer\\panCancer\\13.cliCor")                     #修改工作目录
file="singleGeneExp.txt"                                                      #输入文件
rt=read.table(file,sep="\t",header=T,check.names=F,row.names=1)              #读取表达数据文件
cli1=read.table("clinical.txt",sep="\t",header=T,check.names=F,row.names=1) 
cli<-select(cli1,4)
cli<-filter(cli,!(cli[,1]==''))
#读取临床数据文件
gene=colnames(rt)[1]

clinical=colnames(cli)[1]
#对肿瘤类型循环，进行临床相关性分析
outTab=data.frame()
for(i in levels(as.factor(rt$CancerType))){
	rt1=rt[(rt[,"CancerType"]==i),]
	#交集样品
	data=cbind(rt1,gene=rt1[,gene])
	data=as.matrix(data[,c(gene,"gene")])
	if(nchar(row.names(data)[1])!=nchar(row.names(cli)[1])){
		row.names(data)=gsub(".$","",row.names(data))}
	data=avereps(data)
	sameSample=intersect(row.names(data),row.names(cli))
	sameData=data[sameSample,]
	sameClinical=cli[sameSample,]
	cliExpData=cbind(as.data.frame(sameClinical),sameData)
	if(nrow(cliExpData)==0){next}
	#设置比较组
	group=levels(factor(cliExpData$sameClinical))
	comp=combn(group,2)
	my_comparisons=list()
    for(j in 1:ncol(comp)){my_comparisons[[j]]<-comp[,j]}
	#绘制boxplot
	boxplot=ggboxplot(cliExpData, x="sameClinical", y="gene", color="sameClinical",
	          xlab=clinical,
	          ylab=paste(gene,"expression"),
	          legend.title=clinical,
	          title=paste0("Cancer: ",i),
	          add = "jitter")+ 
	stat_compare_means(comparisons = my_comparisons)
	pdf(file=paste0(clinical,".",i,".pdf"),width=5.5,height=5)
	print(boxplot)
	dev.off()
}