R可视化：使用EnhancedVolcano包可视化火山图

使用EnhancedVolcano包可视化火山图

火山图是用来展示差异表达的基因或者物种，其常常出现RNA-seq，amplicon-seq分析的结果中。其主要根据差异分析结果的p值和Fold change值进行展示结果。更多知识分享请到 https://zouhua.top/。

介绍

标准的火山图常用来展示差异分析的结果，它具有两个指标：

• 差异表达的显著性值：p值或基于p值得adjust-P（差异分析过程涉及到多重校验，因此要对多次检验的p进行校正，校正的方法一般有FDR, BH等等）；
• 差异表达程度值：Fold change值也即使倍数变化值；

数据

Deseq2的结果数据。

• 箭头指向的红框是组间比较，根据Fold changes值设置富集方向，并在图中表示出来；
• DataFrame的红框部分是火山图的坐标轴，FoldChange是x轴，padj是y轴（可以对其取）；

实操

• 读取数据：

  library(dplyr)
  library(tibble)
  library(ggplot2)
  library(EnhancedVolcano)
  
  # pre-run : save(miRNA_Deseq2, file="../../Result/Differential_expression/RNA_DEseq2.RData")
  
  load("../../Result/Differential_expression/RNA_DEseq2.RData")
  

• 画图函数：

  • 设置富集方向标签；

  • 设置差异RNA的名称；

    Volcanofun <- function(prof, type="miRNA", FDR=0.05, lgFC=2){
      # prof <- miRNA.des2
      # type <- "miRNA"
      # FDR <- 0.05
      # lgFC <- 2
    
      geneid2symbolfun <- function(diff_data){
        require(biomaRt)
        require(curl)
        mart <- useDataset("hsapiens_gene_ensembl", useMart("ensembl"))
        # origin 
        ensembl_gene_id <- rownames(diff_data)
        if(!file.exists("../../Result/Differential_expression/geneid2symbol.csv")){
          gene_symbols <- getBM(attributes=c('ensembl_gene_id','external_gene_name',
                                             "transcript_biotype","description"),
                                              filters='ensembl_gene_id',
                                              values=ensembl_gene_id, mart = mart) 
          write.csv(gene_symbols, "../../Result/Differential_expression/geneid2symbol.csv", row.names = F)
        }else{
          gene_symbols <- fread("../../Result/Differential_expression/geneid2symbol.csv")
        }
        
        return(gene_symbols)
      }
        
      if(type == "miRNA"){
        dat <- prof
        label <- rownames(dat)
      }else{
        gene_symbols <- geneid2symbolfun(prof)
        dat_cln <- gene_symbols %>% filter(transcript_biotype == type)
        dat <- subset(prof, rownames(prof)%in%dat_cln$ensembl_gene_id)
        label <- NA
      }
      
      # over- or under- represent RNA
      keyvals <- ifelse(
        dat$log2FoldChange < -lgFC, 'royalblue',
          ifelse(dat$log2FoldChange > lgFC, 'gold',
            'black'))
      names(keyvals)[keyvals == 'gold'] <- 'high'
      names(keyvals)[keyvals == 'black'] <- 'mid'
      names(keyvals)[keyvals == 'royalblue'] <- 'low'
      
      pl <- EnhancedVolcano(dat,
        lab = label,                                  # 显示差异RNA
        x = 'log2FoldChange',                         
        y = 'padj',
        xlim = c(-4, 4),
        ylim = c(-10, 50),
        xlab = bquote(~Log[2]~ 'fold change'),
        ylab = bquote(~-Log[10]~ adjusted~italic(P)),
        title = paste('RB versus PB', "in", type),
        pCutoff = FDR,                                
        FCcutoff = lgFC,
        pointSize = 2.0,
        labSize = 4.0,
        col = c('black', 'black', 'black', 'red3'),
        shape = 16,
        colAlpha = 1,
        drawConnectors = TRUE,
        legendLabSize = 16,
        legendIconSize = 5.0,
        selectLab = rownames(dat)[which(names(keyvals) %in% c('high', 'low'))],
        colCustom = keyvals)+
        scale_x_continuous(breaks = seq(-4, 4, 1),
                           limits = c(-4.5, 4.5))+
        scale_y_continuous(breaks = seq(-5, 40, 5),
                           limits = c(-6, 46))+
        theme_bw()+
        theme(plot.title = element_text(hjust = .5, size = 14))
      
      return(pl)
    }
    

  • Run结果：

    # miRNA
    Volcanofun(miRNA.des2, type="miRNA")
    
    # lncRNA
    Volcanofun(RNA.des2, type="lncRNA")
    

参考

1. EnhancedVolcano；
2. Color points by label
3. 火山图含义

参考文章如引起任何侵权问题，可以与我联系，谢谢。