CODEWITHSUNDEEP
rggplot2geom-raster
bgene
bgene

Reputation: 21

Dual color rectangles from ggplot2 geom_raster()

My question is similar to this link. In this example, 1 is coded for mutation, 0 for wildtype and NA for not available. My dataframe is set up identically, however it may contain two or more types of mutation per gene. I would like to generate a similar figure, except in cases where a gene has two types of mutations, I would like the square to be cut in half and both types of mutations to be colored in, similar to this example. Currently if a gene in a subject has two mutations, the second mutation fill over writes the first. Thank you in advance for taking out the time to help.

dat <- expand.grid(gene=1:10, subj=1:50)
dat$mut <- as.factor(sample(c(rep(0,300),rep(1,200)),500))
dat$mut[sample(500,300)] <- NA
dat[501,] = c(10,50,1) #included from comment below
ggplot(dat, aes(x=subj, y=gene, fill=mut)) +
  geom_raster() +
  scale_fill_manual(values = c("#8D1E0B","#323D8D"), na.value="#FFFFFF") +
  scale_x_discrete("Subject") +
  scale_y_continuous(breaks=1:10,
    labels=c("D0","D1","D2","D3","D4","D5","D6","D7","D8","D9")) +
  guides(fill=FALSE) +
  theme(
    axis.ticks.x=element_blank(), axis.ticks.y=element_blank(),
    axis.text.x=element_blank(), axis.text.y=element_text(colour="#000000"), 
    axis.title.x=element_text(face="bold"), axis.title.y=element_blank(),
    panel.grid.major.x=element_blank(), panel.grid.major.y=element_blank(),
    panel.grid.minor.x=element_blank(), panel.grid.minor.y=element_blank(), 
    panel.background=element_rect(fill="#ffffff")
  )

enter image description here

Upvotes: 1

Views: 620

Answers (1)

Athanasia Mowinckel
Athanasia Mowinckel

Reputation: 462

I can't see that your data ever has any multiple entries for a subject and gene? Nothing is being overwritten, because there is nothing to overwrite.

I added a repetition of the last line, but changed mut to 1 to show. I also changed from raster to tile, and altered the opacity, so that tiles with several values would have a different colour.

If you want something like the plot you link to, you need to create shift and height vector too, like that post shows, so each tile is again subsected.

    dat[501,] = c(10,50,1)

    ggplot(dat, aes(x=subj, y=gene)) +
      geom_tile(alpha=.5,aes(fill=mut), show.legend = F) +
      scale_fill_manual(values = c("#8D1E0B","#323D8D"), na.value="transparent") +
      scale_x_discrete("Subject") +
      scale_y_continuous(breaks=1:10,
                         labels=c("D0","D1","D2","D3","D4","D5","D6","D7","D8","D9")) +
      theme(
        axis.ticks.x=element_blank(), axis.ticks.y=element_blank(),
        axis.text.x=element_blank(), axis.text.y=element_text(colour="#000000"), 
        axis.title.x=element_text(face="bold"), axis.title.y=element_blank(),
        panel.grid.major.x=element_blank(), panel.grid.major.y=element_blank(),
        panel.grid.minor.x=element_blank(), panel.grid.minor.y=element_blank(), 
        panel.background=element_rect(fill="#ffffff")
      )

enter image description here

Upvotes: 2

Related Questions