bernatgel
commented
4 years ago Hi Luca,
That functionality is already available in karyoploteR :) There is a kpHeatmap function that can do what you want. Unfortunately I have not yet reached the point of having it added to the karyoploteR tutorial and it's only mentioned briefly in the vignette. So it's understandable that you didn't find it.
In any case, I'll propose you another way to achieve what you are looking for: using simple rectangles (kpRect) and combining them with a function automatically computing their color (colByValue, again not yet added to the tutorial) and a function to atutomatically determine the vertical positioning (autotrack).
Since I don't have access to the actual data in the plot you linked, we'll start creating some random data with regioneR::createRandomRegions
library(karyoploteR)
nsamples <- 40
regs <- lapply(1:nsamples, function(i) {
r <- createRandomRegions(length.mean=5e6, length.sd=5e6, non.overlapping=TRUE)
r$val <- runif(length(r))
return(r)
})After that, we can create our "pseudoheatmap". We'll loop over the samples and for each one we'll plot its regions as rectangles. The color of the rectangle will be determined by its value using
col=colByValue(regs[[i]]$val, colors = c("red", "white", "blue"), min = 0, max = 1)That function will return a vector of colors depending on the value associated to each rectangle, with 0 beign "red", 0.5 "white" and 1 "blue". You can change the colors and specify any color combination you want. That function uses colorRamp internally, so you can easily find more information on how these colors are selected.
In addition, the vertical position of each sample will be computed with autotrack in
r0=autotrack(i, nsamples)So the sample will be positioned at position i out of nsamples (in this case, 40). autotrack is reeeeeeally handy, I recommend you to take a look at the autotrack tutorial.
So the complete code would be something like this
kp <- plotKaryotype(plot.type=4)
for(i in 1:nsamples) {
kpPlotRegions(kp, regs[[i]],
col=colByValue(value = regs[[i]]$val, colors = c("red", "white", "blue"), min = 0, max = 1),
r0=autotrack(i, nsamples))
}And that will create an image like this one
This is a good start and demonstrate how to create heatmaps with karyoploteR, but you can go muuuuuch further than that.
Changing the parameters a bit, adding labels and colored rectangles, chromosome separators, etc... we can change the look and feel of the plot to resemble much more that of the original plot.
pp <- getDefaultPlotParams(4)
pp$bottommargin <- 10
pp$leftmargin <- 0.02
pp$data1inmargin <- 3
pp$topmargin <- 15
kp <- plotKaryotype(plot.type=4, ideogram.plotter = NULL, labels.plotter = NULL, plot.params = pp)
kpDataBackground(kp, col=colByChr(kp$genome, colors = c("#EEEEEE", "#999999")), data.panel = "ideogram")
kpText(kp, chr=kp$chromosomes, x=mid(kp$genome), y=0.5, labels = gsub("chr", "", kp$chromosomes), data.panel = "ideogram")
kpAddChromosomeSeparators(kp)
chrY.len <- kp$chromosome.lengths["chrY"]
for(i in 1:nsamples) {
at <- autotrack(i, nsamples)
kpRect(kp, chr="chrY", x0=chrY.len+25e6, x1=chrY.len+50e6, y0=0.1, y1=0.9, r0=at, col=rainbow(nsamples)[i], border=NA)
kpAbline(kp, h=0, col="lightgray", r0=at)
kpAddLabels(kp, labels = paste0("S", i), side = "right", r0=at, cex=1.3, label.margin = 0.02)
kpPlotRegions(kp, regs[[i]],
col=colByValue(value = regs[[i]]$val, colors = c("red", "white", "blue"), min = 0, max = 1),
r0=at)
}And you will get this image
Hope this helps! If you have any other request feel free to ask :)
Oh, and if you are plotting copy number data take a look at our companion package CopyNumberPlots (github repo: https://github.com/bernatgel/CopyNumberPlots) for some handy functions to plot copy-number related data!
lbeltrame
Example of what I mean:
https://clincancerres.aacrjournals.org/content/clincanres/23/9/2223/F1.large.jpg?width=800&height=600&carousel=1
it's useful, IMO, not only for quantitative data, but also if you want to plot discrete data over the genome (yes/no).