阅读 332

差异分析+火山图+COX模型构建

简书：没有生物学重复的转录组数据怎么进行差异分析？

edgeR需要的数据是reads数，可以设置BCV值，做单样本的差异分析。
edgeR包可以做无重复的差异分析，不过需要认为指定一个dispersion值(设置BCV值)，这样得到的结果比较主观，不同的人就可以有不同的结果。通常如果是实验控制的好的人类数据，那么选择BCV=0.4，比较好的模式生物选择BCV=0.1。

6 LUAD验证

6.2 差异基因分析和功能分析

差异基因分析

library(limma)
rt=read.table("OV.txt",header=T,sep="\t",check.names=F)
#如果一个基因占了多行，取均值
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<-as.data.frame(data)
nrow(data)
data=data[rowMeans(data)>0.5,]

group_list=c(rep("high",177),rep("low",176))  
table(group_list)



library(edgeR)

dge <- DGEList(counts=data,group=group_list)
dge$samples$lib.size <- colSums(dge$counts)
dge <- calcNormFactors(dge) 

design <- model.matrix(~0+group_list)
rownames(design)<-colnames(dge)
colnames(design)<-levels(group_list)

dge <- estimateGLMCommonDisp(dge,design)
dge <- estimateGLMTrendedDisp(dge, design)
dge <- estimateGLMTagwiseDisp(dge, design)

fit <- glmFit(dge, design)
fit2 <- glmLRT(fit, contrast=c(-1,1)) 

DEG=topTags(fit2, n=nrow(data))
DEG=as.data.frame(DEG)
#logFC_cutoff <- with(DEG,mean(abs(logFC)) + 2*sd(abs(logFC)) )
logFC_cutoff <- 2
DEG$change = as.factor(
  ifelse(DEG$PValue < 0.05 & abs(DEG$logFC) > logFC_cutoff,
         ifelse(DEG$logFC > logFC_cutoff ,'UP','DOWN'),'NOT'))
head(DEG)
table(DEG$change)
edgeR_DEG <- DEG

#提取差异基因表达矩阵
DEG<-DEG[DEG$change!="NOT",]
library(data.table)
#fintersect求交集
a<-data.table(rownames(DEG))
diff<-data[match(a$V1,row.names(data)),]
write.csv(diff,file = "diff.csv",quote=F,row.names = T)

source("3-plotfunction.R")
logFC_cutoff <- 2
dat = log(data+1)
pca.plot = draw_pca(dat,group_list)

cg2 = rownames(edgeR_DEG)[edgeR_DEG$change !="NOT"]
h2 = draw_heatmap(data[cg2,],group_list)
v2 = draw_volcano(test = edgeR_DEG[,c(1,4,6)],pkg = 2)
v2
library(patchwork)
(h2) / (v2) +plot_layout(guides = 'collect')
ggsave("heat_volcano.png",width = 7,height = 5)
火山图

功能分析

library("org.Hs.eg.db")
rt=read.csv("diff.csv",sep=",",check.names=F,header=T)
rt<-rt[,1]
rt<-as.data.frame(rt)

genes=as.vector(rt[,1])
entrezIDs <- mget(genes, org.Hs.egSYMBOL2EG, ifnotfound=NA)
entrezIDs <- as.character(entrezIDs)
out=cbind(rt,entrezID=entrezIDs)
write.table(out,file="id.txt",sep="\t",quote=F,row.names=F)


###############GO分析
library("clusterProfiler")
library("org.Hs.eg.db")
library("enrichplot")
library("ggplot2")

rt=read.table("id.txt",sep="\t",header=T,check.names=F)
rt=rt[is.na(rt[,"entrezID"])==F,]

cor=rt$cor
gene=rt$entrezID
names(cor)=gene

#GO富集分析
kk <- enrichGO(gene = gene,
               OrgDb = org.Hs.eg.db, 
               pvalueCutoff =0.05, 
               qvalueCutoff = 0.05,
               ont="all",
               readable =T)
write.table(kk,file="GO.txt",sep="\t",quote=F,row.names = F)

#柱状图
tiff(file="GO.tiff",width = 26,height = 20,units ="cm",compression="lzw",bg="white",res=600)
barplot(kk, drop = TRUE, showCategory =10,split="ONTOLOGY") + facet_grid(ONTOLOGY~., scale='free')
dev.off()


#####################KEGG分析
library("clusterProfiler")
library("org.Hs.eg.db")
library("enrichplot")
library("ggplot2")

setwd("C:\\Users\\lexb4\\Desktop\\CCLE\\09.KEGG")
rt=read.table("id.txt",sep="\t",header=T,check.names=F)
rt=rt[is.na(rt[,"entrezID"])==F,]

cor=rt$cor
gene=rt$entrezID
names(cor)=gene

#kegg富集分析
kk <- enrichKEGG(gene = gene, organism = "hsa", pvalueCutoff =0.05, qvalueCutoff =0.05)
write.table(kk,file="KEGG.txt",sep="\t",quote=F,row.names = F)

#柱状图
tiff(file="barplot.tiff",width = 20,height = 12,units ="cm",compression="lzw",bg="white",res=600)
barplot(kk, drop = TRUE, showCategory = 20)
dev.off()

6.3 批量KM曲线

输入数据KM.TXT，数据是FPKM.


KM.txt
library(survival)
rt=read.table("KM.txt",header=T,sep="\t",check.names=F)
outTab=data.frame()

for(gene in colnames(rt[,4:ncol(rt)])){
  a=rt[,gene]<=median(rt[,gene])
  diff=survdiff(Surv(futime, fustat) ~a,data = rt)
  pValue=1-pchisq(diff$chisq,df=1)
  outTab=rbind(outTab,cbind(gene=gene,pvalue=pValue))
  
  fit <- survfit(Surv(futime, fustat) ~ a, data = rt)
  summary(fit)
  #只将p<0.05的基因画出来。
  if(pValue<0.05){
    if(pValue<0.001){
      pValue="<0.001"
    }else{
      pValue=round(pValue,3)
      pValue=paste0("=",pValue)
    }
    pdf(file=paste(gene,".survival.pdf",sep=""),
        width=6,
        height=6)
    plot(fit, 
         lwd=2,
         col=c("red","blue"),
         xlab="Time (year)",
         mark.time=T, #显示删失数据
         ylab="Survival rate",
         main=paste(gene,"(p", pValue ,")",sep=""))
    legend("topright", 
           c("High","Low"), 
           lwd=2, 
           col=c("red","blue"))
    dev.off()
  }
}
write.table(outTab,file="survival.xls",sep="\t",row.names=F,quote=F)

6.4 差异基因进行COX模型构建

用FPKM数据做COX模型的构建，用OS相关的基因做分析。
输入数据exptime.txt

exptime.txt

单因素cox

pFilter=0.01                            #定义单因素显著性
library(survival)                                                  #引用包
rt=read.table("expTime.txt",header=T,sep="\t",check.names=F,row.names=1)       #读取输入文件

sigGenes=c("futime","fustat")
outTab=data.frame()
for(i in colnames(rt[,3:ncol(rt)])){
 cox <- coxph(Surv(futime, fustat) ~ rt[,i], data = rt)
 coxSummary = summary(cox)
 coxP=coxSummary$coefficients[,"Pr(>|z|)"]
 outTab=rbind(outTab,
              cbind(id=i,
              HR=coxSummary$conf.int[,"exp(coef)"],
              HR.95L=coxSummary$conf.int[,"lower .95"],
              HR.95H=coxSummary$conf.int[,"upper .95"],
              pvalue=coxSummary$coefficients[,"Pr(>|z|)"])
              )
  if(coxP<pFilter){
      sigGenes=c(sigGenes,i)
  }
}
write.table(outTab,file="uniCox.xls",sep="\t",row.names=F,quote=F)
uniSigExp=rt[,sigGenes]
uniSigExp=cbind(id=row.names(uniSigExp),uniSigExp)
write.table(uniSigExp,file="uniSigExp.txt",sep="\t",row.names=F,quote=F)

LASSO回归

library("glmnet")
library("survival")
rt=read.table("uniSigExp.txt",header=T,sep="\t",row.names=1,check.names=F)     #读取文件
rt$futime[rt$futime<=0]=1

x=as.matrix(rt[,c(3:ncol(rt))])
y=data.matrix(Surv(rt$futime,rt$fustat))

fit <- glmnet(x, y, family = "cox", maxit = 1000)
pdf("lambda.pdf")
plot(fit, xvar = "lambda", label = TRUE)
dev.off()

cvfit <- cv.glmnet(x, y, family="cox", maxit = 1000)
pdf("cvfit.pdf")
plot(cvfit)
abline(v=log(c(cvfit$lambda.min,cvfit$lambda.1se)),lty="dashed")
dev.off()

coef <- coef(fit, s = cvfit$lambda.min)
index <- which(coef != 0)
actCoef <- coef[index]
lassoGene=row.names(coef)[index]
lassoGene=c("futime","fustat",lassoGene)
lassoSigExp=rt[,lassoGene]
lassoSigExp=cbind(id=row.names(lassoSigExp),lassoSigExp)
write.table(lassoSigExp,file="lassoSigExp.txt",sep="\t",row.names=F,quote=F)

多因素COX模型的构建

library(survival)
library(survminer)

rt=read.table("lassoSigExp.txt",header=T,sep="\t",check.names=F,row.names=1)  #读取train输入文件
rt[,"futime"]=rt[,"futime"]/365

#使用train组构建COX模型
multiCox=coxph(Surv(futime, fustat) ~ ., data = rt)
multiCox=step(multiCox,direction = "both")
multiCoxSum=summary(multiCox)

#输出模型相关信息
outTab=data.frame()
outTab=cbind(
             coef=multiCoxSum$coefficients[,"coef"],
             HR=multiCoxSum$conf.int[,"exp(coef)"],
             HR.95L=multiCoxSum$conf.int[,"lower .95"],
             HR.95H=multiCoxSum$conf.int[,"upper .95"],
             pvalue=multiCoxSum$coefficients[,"Pr(>|z|)"])
outTab=cbind(id=row.names(outTab),outTab)
write.table(outTab,file="multiCox.xls",sep="\t",row.names=F,quote=F)

#绘制森林图
pdf(file="forest.pdf",
       width = 8,             #图片的宽度
       height = 5,            #图片的高度
       )
ggforest(multiCox,
         main = "Hazard ratio",
         cpositions = c(0.02,0.22, 0.4), 
         fontsize = 0.7, 
         refLabel = "reference", 
         noDigits = 2)
dev.off()

#输出train组风险文件
riskScore=predict(multiCox,type="risk",newdata=rt)           #利用train得到模型预测train样品风险
coxGene=rownames(multiCoxSum$coefficients)
coxGene=gsub("`","",coxGene)
outCol=c("futime","fustat",coxGene)
medianTrainRisk=median(riskScore)
risk=as.vector(ifelse(riskScore>medianTrainRisk,"high","low"))
write.table(cbind(id=rownames(cbind(rt[,outCol],riskScore,risk)),cbind(rt[,outCol],riskScore,risk)),
    file="riskTrain.txt",
    sep="\t",
    quote=F,
    row.names=F)
####################下面是验证集的风险文件输出#################
#输出test组风险文件
rtTest=read.table("test.txt",header=T,sep="\t",check.names=F,row.names=1)          #读取test输入文件
rtTest[,"futime"]=rtTest[,"futime"]/365
riskScoreTest=predict(multiCox,type="risk",newdata=rtTest)      #利用train得到模型预测test样品风险
riskTest=as.vector(ifelse(riskScoreTest>medianTrainRisk,"high","low"))
write.table(cbind(id=rownames(cbind(rtTest[,outCol],riskScoreTest,riskTest)),cbind(rtTest[,outCol],riskScore=riskScoreTest,risk=riskTest)),
    file="riskTest.txt",
    sep="\t",
    quote=F,
    row.names=F)

6.5 风险图形绘制

风险生存曲线

library(survival)

#绘制train组生存曲线
rt=read.table("riskTrain.txt",header=T,sep="\t")
diff=survdiff(Surv(futime, fustat) ~risk,data = rt)
pValue=1-pchisq(diff$chisq,df=1)
pValue=signif(pValue,4)
pValue=format(pValue, scientific = TRUE)
fit <- survfit(Surv(futime, fustat) ~ risk, data = rt)
summary(fit)    #查看五年生存率
pdf(file="survivalTrain.pdf",width=5.5,height=5)
plot(fit, 
     lwd=2,
     col=c("red","blue"),
     xlab="Time (year)",
     ylab="Survival rate",
     main=paste("Survival curve (p=", pValue ,")",sep=""),
     mark.time=T)
legend("topright", 
       c("high risk", "low risk"),
       lwd=2,
       col=c("red","blue"))
dev.off()

#绘制test组生存曲线
rt=read.table("riskTest.txt",header=T,sep="\t")
diff=survdiff(Surv(futime, fustat) ~risk,data = rt)
pValue=1-pchisq(diff$chisq,df=1)
pValue=signif(pValue,4)
pValue=format(pValue, scientific = TRUE)
fit <- survfit(Surv(futime, fustat) ~ risk, data = rt)
summary(fit)    #查看五年生存率
pdf(file="survivalTest.pdf",width=5.5,height=5)
plot(fit, 
     lwd=2,
     col=c("red","blue"),
     xlab="Time (year)",
     ylab="Survival rate",
     main=paste("Survival curve (p=", pValue ,")",sep=""),
     mark.time=T)
legend("topright", 
       c("high risk", "low risk"),
       lwd=2,
       col=c("red","blue"))
dev.off()
KM曲线

风险ROC曲线

library(survivalROC)

rt=read.table("riskTrain.txt",header=T,sep="\t",check.names=F,row.names=1)
pdf(file="rocTrain.pdf",width=6,height=6)
par(oma=c(0.5,1,0,1),font.lab=1.5,font.axis=1.5)
roc=survivalROC(Stime=rt$futime, status=rt$fustat, marker = rt$riskScore, 
      predict.time =1, method="KM")
plot(roc$FP, roc$TP, type="l", xlim=c(0,1), ylim=c(0,1),col='red', 
  xlab="False positive rate", ylab="True positive rate",
  main=paste("ROC curve (", "AUC = ",round(roc$AUC,3),")"),
  lwd = 2, cex.main=1.3, cex.lab=1.2, cex.axis=1.2, font=1.2)
abline(0,1)
dev.off()

rt=read.table("riskTest.txt",header=T,sep="\t",check.names=F,row.names=1)
pdf(file="rocTest.pdf",width=6,height=6)
par(oma=c(0.5,1,0,1),font.lab=1.5,font.axis=1.5)
roc=survivalROC(Stime=rt$futime, status=rt$fustat, marker = rt$riskScore, 
      predict.time =1, method="KM")
plot(roc$FP, roc$TP, type="l", xlim=c(0,1), ylim=c(0,1),col='red', 
  xlab="False positive rate", ylab="True positive rate",
  main=paste("ROC curve (", "AUC = ",round(roc$AUC,3),")"),
  lwd = 2, cex.main=1.3, cex.lab=1.2, cex.axis=1.2, font=1.2)
abline(0,1)
dev.off()
ROC曲线

作者:医学小白学生信

原文链接:https://www.jianshu.com/p/9dec2a5214cd

文章分类
后端
文章标签
差异
cox
模型
版权声明：本站是系统测试站点，无实际运营。本文内容由互联网用户自发贡献，该文观点仅代表作者本人。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容， 请发送邮件至 XXXXXXo@163.com 举报，一经查实，本站将立刻删除。
相关推荐