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

#install.packages("ggpubr")


#引用包
library(reshape2)
library(ggpubr)
library(limma)
library(GSEABase)
library(GSVA)

expFile="GSE19151.normalize.txt"            #表达数据文件
geneFile="interGenes.txt"                 #基因列表文件
gmtFile="c5.go.Hs.symbols.gmt"     #基因集文件
setwd("F:\\血栓\\11.GSVA分析")     #设置工作目录

#读取表达输入文件,并对输入文件整理
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)

#去除对照组的样品
Type=gsub("(.*)\\_(.*)", "\\2", colnames(data))
data=data[,Type=="Treat",drop=F]

#读取基因集文件
geneSets=getGmt(gmtFile, geneIdType=SymbolIdentifier())

#GSVA分析
ssgseaPar <- ssgseaParam(data, geneSets, minSize=5, maxSize=500)
gsvaScore <- gsva(ssgseaPar)
#对打分进行矫正
normalize=function(x){
  return((x-min(x))/(max(x)-min(x)))}
gsvaScore=normalize(gsvaScore)
gsvaScore=gsvaScore[apply(gsvaScore,1,sd)>0.01,]

#读取基因列表文件
geneRT=read.table(geneFile, header=F, sep="\t", check.names=F)

#对基因进行循环
for(gene in as.vector(geneRT[,1])){
	#根据目标基因的表达量对样品进行分组
	lowName=colnames(data)[data[gene,]<median(data[gene,])]       #低表达组的样品
	highName=colnames(data)[data[gene,]>=median(data[gene,])]     #高表达组的样品
	lowScore=gsvaScore[,lowName]
	highScore=gsvaScore[,highName]
	scoreData=cbind(lowScore, highScore)
	conNum=ncol(lowScore)
	treatNum=ncol(highScore)
	Type=c(rep("Control",conNum), rep("Treat",treatNum))

	#对通路进行循环, 进行通路的差异分析
	outTab=data.frame()
	for(i in row.names(scoreData)){
		test=t.test(scoreData[i,] ~ Type)
		t=test$statistic
		pvalue=test$p.value
		if(test$estimate[2]>test$estimate[1]){t=abs(t)}else{t=-abs(t)}
		Sig=ifelse(pvalue>0.05, "Not", ifelse(t>0,"Up","Down"))
		outTab=rbind(outTab, cbind(Pathway=i, t, pvalue, Sig))
	}
	notSigTab=outTab[outTab$Sig=="Not",]
	notSigTab=notSigTab[order(as.numeric(notSigTab$t)),]
	sigTab=outTab[outTab$Sig!="Not",]
	sigTab=sigTab[order(as.numeric(sigTab$t)),]
	if(nrow(sigTab)>20){
		outTab=rbind(sigTab[c(1:10,((nrow(sigTab)-9):nrow(sigTab))),],notSigTab[c(1:5,((nrow(notSigTab)-4):nrow(notSigTab))),])
	}else{
		outTab=rbind(sigTab,notSigTab[c(1:5,((nrow(notSigTab)-4):nrow(notSigTab))),])
	}

	#绘制柱状图
	pdf(file=paste0(gene,".barplot.pdf"), width=13, height=7)
	outTab$t=as.numeric(outTab$t)
	outTab$Sig=factor(outTab$Sig, levels=c("Down", "Not", "Up"))
	gg1=ggbarplot(outTab, x="Pathway", y="t", fill = "Sig", color = "white",
			palette=c("#74C476FF","grey","#E6550DFF"), sort.val = "asc", sort.by.groups = T,
			rotate=TRUE, title=gene, legend.title="Group", legend="right",
			xlab="", ylab="t value of GSVA score",  x.text.angle=60)
	print(gg1)
	dev.off()
}