火山图可以方便直观地展示两个样本间基因差异表达的分布情况。通常横坐标用log2（fold change）表示，差异越大的基因分布在两端，纵坐标用-log10（pvalue）表示，T检验显著性P值的负对数。通常差异倍数越大的基因T检验越显著，所以往往关注左上角和右上角的值。我们根据计算结果在图上按照FDR和我们自定义的logFC值来将这部分显著变化的基因用不同颜色标示出来以区分，这类图像往往呈现类似火山爆发的样子，于是就被叫做“火山图”（volcano plot）了。
library(ggplot2) diff <- read.csv("DEG.Out.csv",header = T) logFC <-diff$logFC adj <- diff$FDR data <- data.frame(logFC=logFC,padj=padj) data$sig[(data$padj > 0.05|data$padj=="NA")|(data$logFC < 0.5)& data$logFC > -0.5] <- "no" data$sig[data$padj <= 0.05 & data$logFC >= 0.5] <- "up" data$sig[data$padj <= 0.05 & data$logFC <= -0.5] <- "down" # 选最大值作为xlim的上下边界 x_lim <- max(logFC,-logFC) # 绘制火山图 library(ggplot2) library(RColorBrewer) pdf(file = "miRNA_volcano.pdf",width=8,height=8) theme_set(theme_bw()) p <- ggplot(data,aes(logFC,-1*log10(padj), color = sig))+geom_point()+ xlim(-5,5) + labs(x="log2(FoldChange)",y="-log10(FDR)") p <- p + scale_color_manual(values =c("#0072B5","grey","#BC3C28"))+ geom_hline(yintercept=-log10(0.05),linetype=4)+ geom_vline(xintercept=c(-0.5,0.5),linetype=4) p <- p +theme(panel.grid =element_blank())+ theme(axis.line = element_line(size=0))+ylim(0,15) p <- p +guides(colour = FALSE) p <- p +theme(axis.text=element_text(size=20),axis.title=element_text(size=20)) p dev.off() print(p)
