TCGA差异分析及热图火山图绘制

输入文件：

immLncExp.png

#引用包
library(limma)
library(pheatmap)
expFile="immLncExp.txt"     #表达输入文件
fdrFilter=0.05              #fdr临界值
logFCfilter=1               #logFC临界值
setwd("E:\\Master research")       #设置工作目录

#读取输入文件
rt=read.table(expFile, header=T, sep="\t", check.names=F)
rt=as.matrix(rt)
rownames(rt)=rt[,1]
exp=rt[,2:ncol(rt)]
dimnames=list(rownames(exp),colnames(exp))
data=matrix(as.numeric(as.matrix(exp)),nrow=nrow(exp),dimnames=dimnames)
data=avereps(data)
data=data[rowMeans(data)>0,]

#正常和肿瘤数目
group=sapply(strsplit(colnames(data),"\\-"),"[",4)
group=sapply(strsplit(group,""),"[",1)
group=gsub("2", "1", group)
conNum=length(group[group==1])       #正常组样品数目
treatNum=length(group[group==0])     #肿瘤组样品数目
grade=c(rep(1,conNum), rep(2,treatNum))

#差异分析
outTab=data.frame()
for(i in row.names(data)){
    geneName=unlist(strsplit(i,"\\|",))[1]
    geneName=gsub("\\/", "_", geneName)
    rt=rbind(expression=data[i,], grade=grade)
    rt=as.matrix(t(rt))
    wilcoxTest=wilcox.test(expression ~ grade, data=rt)
    conGeneMeans=mean(data[i,1:conNum])
    treatGeneMeans=mean(data[i,(conNum+1):ncol(data)])
    logFC=log2(treatGeneMeans)-log2(conGeneMeans)
    pvalue=wilcoxTest$p.value
    conMed=median(data[i,1:conNum])
    treatMed=median(data[i,(conNum+1):ncol(data)])
    diffMed=treatMed-conMed
    if( ((logFC>0) & (diffMed>0)) | ((logFC<0) & (diffMed<0)) ){  
          outTab=rbind(outTab,cbind(gene=i,conMean=conGeneMeans,treatMean=treatGeneMeans,logFC=logFC,pValue=pvalue))
     }
}
pValue=outTab[,"pValue"]
fdr=p.adjust(as.numeric(as.vector(pValue)), method="fdr")
outTab=cbind(outTab, fdr=fdr)

#输出所有基因的差异情况
write.table(outTab,file="all.xls",sep="\t",row.names=F,quote=F)

#输出差异表格
outDiff=outTab[( abs(as.numeric(as.vector(outTab$logFC)))>logFCfilter & as.numeric(as.vector(outTab$fdr))(2*geneNum)){
    hmGene=diffGeneName[c(1:geneNum,(diffLength-geneNum+1):diffLength)]
}else{
    hmGene=diffGeneName
}
hmExp=log2(data[hmGene,]+0.01)
Type=c(rep("Normal",conNum),rep("Tumor",treatNum))
names(Type)=colnames(data)
Type=as.data.frame(Type)
pdf(file="heatmap.pdf", height=7, width=10)
pheatmap(hmExp, 
         annotation=Type, 
         color = colorRampPalette(c(rep("blue",2), "white", rep("red",2)))(50),
         cluster_cols =F,
         show_colnames = F,
         scale="row",
         fontsize = 10,
         fontsize_row=5,
         fontsize_col=10)
dev.off()

#绘制火山图
pdf(file="vol.pdf",height=5,width=5)
xMax=8
yMax=max(-log10(outTab$fdr))+1
plot(as.numeric(as.vector(outTab$logFC)), -log10(outTab$fdr), xlab="logFC",ylab="-log10(fdr)",
     main="Volcano", ylim=c(0,yMax),xlim=c(-xMax,xMax),yaxs="i",pch=20, cex=0.8)
diffSub=subset(outTab, fdrlogFCfilter)
points(as.numeric(as.vector(diffSub$logFC)), -log10(diffSub$fdr), pch=20, col="red",cex=1)
diffSub=subset(outTab, fdr