This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to cluster the annotation in the heat map?
library(ComplexHeatmap)

set.seed(123)
mat = matrix(rnorm(100), 10)
foo = sample(letters[1:2], 10, replace = TRUE)
colnames(mat) = sapply(letters[1:10], function(x) paste(rep(x, 10), collapse = ""))
cn = colnames(mat)
ha2 = HeatmapAnnotation(foo = foo,
                cn = anno_text(cn, rot = 90, offset = unit(1, "npc") - unit(2, "mm"), just = "right"),
                annotation_height = unit.c(unit(5, "mm"), max_text_width(cn)))
Heatmap(mat, show_column_names = FALSE, bottom_annotation = ha2)

This gave me a heat map like following:

https://imgur.com/CHQdXyb

Here you see in the bottom annotation both a and b are not clustered. I would like to know how can I do each group (color) get clustered. I mean all samples which come under "b" at one side and samples under "a" on one side.

Can anyone please tell me how to do this? Thanks !!

r heatmap clustering rna-seq
• 6,217 views • link
updated by Kevin Blighe Kevin Blighe • written by Biologist Biologist

1 answer

ComplexHeatmap doesn't support splitting that way - it only supports splitting by rows. If you want to split by columns, you could just generate two separate heatmaps and plot them side by side with the draw function draw(hmap1 + hmap2).

In your case, you could also just plot the transpose of your data-matrix and then split by rows, as follows (NB - the key parameter to use is split):

library(ComplexHeatmap)

set.seed(123)

mat=matrix(rnorm(100), 10)

foo=sample(letters[1:2], 10, replace=TRUE)

colnames(mat)=sapply(letters[1:10], function(x) paste(rep(x, 10), collapse=""))

cn = colnames(mat)

ha2 = rowAnnotation(foo=foo,
                #cn=anno_text(cn, rot=0, just="right", which="row"),
                annotation_height=unit.c(unit(5, "mm"), max_text_width(cn)))

draw(Heatmap(t(mat), split=foo, show_row_names=TRUE) + ha2, heatmap_legend_side="left", annotation_legend_side="left")

hmap

• 0 views • link
written by Kevin Blighe Kevin Blighe

Log in to answer this question.

More posts like this