Plot a histogram together with an UpSet plot in R

Question

I have a data frame that looks like this,

  df <- data.frame(type=c("SNP","DEL","SNP","SNP"),geneA=c(1,1,1,0), geneB=c(0,0,1,1), geneC=c(1,0,0,1))

   type geneA geneB geneC
1  SNP     1     0     1
2  DEL     1     0     0
3  SNP     1     1     0
4  SNP     0     1     1

I want to make an UpSet plot in R to find the common genes and
I want to plot the distribution of types (SNP or DEL) in an histogram.

this is my code so far

  upset(df, 
        attribute.plots = list(gridrows=50,
                           plots=list(list(plot=histogram,
                                           x="type"))))

and this is my error, which I can not solve

Error: StatBin requires a continuous x variable: the x variable is discrete.Perhaps you want stat="count"?

ANY help is highly appreciated

If you want to combine ggplot2 components with upset I can suggest an alternative I created for that purpose: https://github.com/krassowski/complex-upset. — krassowski, Aug 03 '21 at 21:30
Its a bit complicate to me how it works. If I try something like this ```upset(df, annotations = list( 'type'=( ggplot(mapping=aes(fill=type)) + geom_bar(stat='count', position='fill')))) ``` I am still getting an error. Maybe because its late at night in here — LDT, Aug 03 '21 at 21:41

score 1 · Accepted Answer · answered Aug 11 '21 at 16:25

I am not sure what the distribution of types (SNP or DEL) in an histogram means exactly. However, because the data of SNP and DEL in your example are binaries, you probably want to plot the counts of SNP and DEL in a bar plot. If that's true, you can try this way:

mybarplot <- function(mydata, x) {
  (ggplot(mydata, aes_string(x = x)) + geom_bar()
  + theme(plot.margin = unit(c(0.5, 0.5, 0, 0), "cm"), 
          legend.key.size = unit(0.4, "cm")))
}

This function takes a data frame and one of its column as its inputs, and generates a barplot as its output. Then, you call this function within your upset function to produce the barplot along with your upset plot.

upset(dftest, attribute.plots = list(gridrows = 50, 
                                     plots = list(list(plot = mybarplot, 
                                                       x = "type"))))

Plot a histogram together with an UpSet plot in R

1 Answers1