Chapter 9 R语言韦恩图

韦恩图主要用来展示不同组数据之间的交集

R语言里画韦恩图的R包最常用的可能是VennDiagram,最多可以展示5组数据之间的交集，每组数据单独准备到一个文件里

library(readxl)

dat01<-read_excel("example_data/09-venndiagram/dat01.xlsx")
dat02<-read_excel("example_data/09-venndiagram/dat02.xlsx")
dat03<-read_excel("example_data/09-venndiagram/dat03.xlsx")
dat04<-read_excel("example_data/09-venndiagram/dat04.xlsx")

library(VennDiagram)

## Loading required package: grid

## Loading required package: futile.logger

## 
## Attaching package: 'VennDiagram'

## The following object is masked from 'package:ggtree':
## 
##     rotate

x<-list(group1=dat01$gene_name,
        group2=dat02$gene_name,
        group3=dat03$gene_name,
        group4=dat04$gene_name)

venn.diagram(x,filename = NULL) -> p1

grid::grid.draw(p1)

有很多参数可以调节，这里介绍一下比较常用的

library(readxl)

dat01<-read_excel("example_data/09-venndiagram/dat01.xlsx")
dat02<-read_excel("example_data/09-venndiagram/dat02.xlsx")
dat03<-read_excel("example_data/09-venndiagram/dat03.xlsx")
dat04<-read_excel("example_data/09-venndiagram/dat04.xlsx")

library(VennDiagram)

x<-list(group1=dat01$gene_name,
        group2=dat02$gene_name,
        group3=dat03$gene_name,
        group4=dat04$gene_name)

venn.diagram(x,
             filename = NULL,
             main="ABC",
             sub="abc",
             fill=c("red","green","blue","yellow"),
             col="grey",
             cat.cex=3) -> p1

grid.draw(p1)

ggplot2没有专门做venn图的函数，有人开发了ggplot2的扩展包用来画韦恩图

ggvenn https://github.com/yanlinlin82/ggvenn 最多能画4组数据

ggvenndiagram https://github.com/gaospecial/ggVennDiagram 最多能画7组数据

这里简单介绍一下ggvenn

安装直接运行install.packages("ggvenn")

准备数据和VennDiagram的一眼

作图代码

library(readxl)

dat01<-read_excel("example_data/09-venndiagram/dat01.xlsx")
dat02<-read_excel("example_data/09-venndiagram/dat02.xlsx")
dat03<-read_excel("example_data/09-venndiagram/dat03.xlsx")
dat04<-read_excel("example_data/09-venndiagram/dat04.xlsx")



x<-list(group1=dat01$gene_name,
        group2=dat02$gene_name,
        group3=dat03$gene_name,
        group4=dat04$gene_name)
library(ggvenn)

## Warning: package 'ggvenn' was built under R version 4.0.5

ggvenn(x)

ggvenn(x,
  fill_color = c("blue", "yellow", "green", "red"),
  fill_alpha = 0.5,
  stroke_color = "grey",
  stroke_alpha = 1,
  stroke_size = 1,
  stroke_linetype = "dashed",
  set_name_color = "red",
  set_name_size = 6,
  text_color = "black",
  text_size = 4,
  label_sep = ",")

VennDiagram 和 ggvenn 的每一个圈展示和每组数据集的多少并不是成比例的，比如第一组10个数据，第二组5个数据，第一组的圈应该比第二组的圈大

如果想要成比例的话，可以使用R包eulerr

这个整理数据集可能会复杂一些

可以画5个以上的韦恩图

dat<-c("First" = 25, 
       "Second" = 5, 
       "Third" = 5,
       "First&Second" = 5, 
       "First&Third" = 5, 
       "Second&Third" = 3,
       "First&Second&Third" = 3)

library(eulerr)

## Warning: package 'eulerr' was built under R version 4.0.5

plot(euler(dat),
     fills = list(fill=c("red","blue",
                         "green","darkgreen",
                         "white","black",
                         "purple"),
                  alpha=0.5),
     quantities = list(c(25,5,5,
                         1,1,1,1),
                       col="black",
                       cex=4),
     labels = list(col="white",font=3,cex=2),
     edges = list(col="darkgreen",lwd=5,
                  lty=1:3),
     main = list(label=c("XiaoMing"),cex=5),
     legend = list(labels=c("A","B","C"),
                   cex=5))

韦恩图如果数据超过5组，还有一种展示方式如下，可以借助UpSetR包实现 https://github.com/hms-dbmi/UpSetR

输入数据和之前介绍的一眼

library(readxl)

dat01<-read_excel("example_data/09-venndiagram/dat01.xlsx")
dat02<-read_excel("example_data/09-venndiagram/dat02.xlsx")
dat03<-read_excel("example_data/09-venndiagram/dat03.xlsx")
dat04<-read_excel("example_data/09-venndiagram/dat04.xlsx")



x<-list(group1=dat01$gene_name,
        group2=dat02$gene_name,
        group3=dat03$gene_name,
        group4=dat04$gene_name)

library(UpSetR)

## Warning: package 'UpSetR' was built under R version 4.0.5

upset(fromList(x), 
      order.by = "freq",
      point.size = 3.5, 
      line.size = 2,
      mainbar.y.label = "AAAA",
      sets.x.label = "BBBB",
      text.scale = c(1.3, 1, 5, 0.75),
      main.bar.color="red")

来自一篇NC论文的实际例子

https://github.com/PlantNutrition/Liyu

library(UpSetR)#R 3.5
otu_RA <- read.delim('example_data/09-venndiagram/otu_RA.txt', header = TRUE, row.names = 1, sep = '\t')
otu_RA[otu_RA > 0] <- 1
p <- upset(otu_RA, 
           nset = 7, 
           nintersects = 10, 
           order.by = c('degree','freq'),
           decreasing = c(TRUE, TRUE),
          mb.ratio = c(0.7, 0.3),
            point.size = 1.8,
            line.size = 1, 
            mainbar.y.label = "Intersection size", 
            sets.x.label = "Set Size", 
            main.bar.color = "#2a83a2", 
          sets.bar.color = "#3b7960",
            queries = list(list(query = intersects, 
                                params = list("BS","RS","RE","VE","SE","LE","P"), 
                                active = T,
                                color="#d66a35", 
                                query.name = "BS vs RS vs RE vs VE vs SE vs LE vs P")))
p