Search code examples
rvenn-diagram

How to add counts to a venneuler venn diagram that was plotted from a matrix

I have several lists of genes (all of various lengths) that I would like to visually compare by venn diagram. I wrote a little function that uses ReShape2 to convert any dataframe of gene names into a matrix of 1's and 0's that Venneuler can use to plot a Venn Diagram. My problem is that I cannot figure out how to extract/calculate the value associated with each part of the venn diagram. Additionally, it'd be nice if I could add those values to my plot in R.

Here's an example of what my data look like:

 A <- c("gene1", "gene2", "gene3", "gene5", "gene12", "", "")
 B <- c("gene1", "gene2", "gene6", "gene7", "", "", "")
 C <- c("gene2", "gene6", "gene7", "gene8", "gene9", "gene13", "gene14")
 D <- c("gene7", "gene8", "gene9", "gene10", "gene11", "gene12", "")
 dat <- data.frame(A,B,C,D)

Function that will convert a table of gene names into a presence/absence matrix that Vennueler can use:

 vennfun <- function(x) { 
  x$id <- seq(1, nrow(x))  #add a column of numbers (required for melt)
  xm <- melt(x, id.vars="id", na.rm=TRUE)  #melt table into two columns (value & variable)
  xc <- dcast(xm, value~variable, fun.aggregate=length)  #remove NA's, list presence/absence of each value for each variable (1 or 0)
  rownames(xc) <- xc$value  #value column = rownames (required for Venneuler)
  xc$value <- NULL  #remove redundent value column
  xc  #output the new dataframe
}

Load required packages:

library(reshape2)
library(venneuler)

Run vennfun and use the output to plot a venn diagram with venneuler:

VennDat <- vennfun(dat)
genes.venn <- venneuler(VennDat)
plot(genes.venn)

My question is: how do I get the number of genes associated with all possible conditions (i.e. A, AB, ABC, ABCD, B, BC, BCD, ABD, ACD, etc.), and/or how do I add these values to my venn diagram?

Thanks!!

Solution

  • If you're willing to change package, you could accomplish this with eulerr (that I am the author of):

    library(eulerr)
genes.venn <- euler(VennDat)
plot(genes.venn, quantities = TRUE)

    enter image description here

    As an aside, this problem does not really lend itself well to a Euler diagram. (The fit is quite poor.) Perhaps you should consider an alternative?