How to have an additional coloumn in a heatmaply heatmap

Question

I am using heatmaply to obtain clustered heatmaps of responses from multiple raters on a series of questions rated using the same Leikert scale (ECOG Performance Status). The heatmap comes out well (though using hierarchical clustering on ordinal data like this may not be best). I would like to display an additional coloumn in the heatmap that has color coded information on an additional variable e.g. Age. The example heatmap I have generated using the package is attached. The coloumn in blue has information on patient's gender but the same is not color coded. I would like to know if the same can be done. Also would welcome any inputs as to the correct clustering methodology to be used for ordinal data.
Original heatmap link

The code used is here:

library(heatmaply)
data4 <- structure(list(UID = c("D1", "D3", "D4", "D5", "D6", "D7", "D8", 
"D9", "D10", "D11", "D12", "D13", "D14", "D15", "D16"), R101 = c(2, 
1, 1, 1, 2, 1, 2, 1, 0, 2, 0, 1, 1, 1, 1), R102 = c(3, 2, 0, 
2, 3, 1, 2, 2, 0, 2, 3, 2, 2, 2, 2), R103 = c(2, 2, 2, 3, 3, 
0, 2, 3, 0, 1, 0, 4, 2, 2, 3), R104 = c(1, 0, 1, 1, 1, 1, 1, 
3, 0, 2, 1, 0, 0, 1, 2), R105 = c(1, 3, 2, 1, 1, 2, 1, 1, 0, 
3, 1, 0, 2, 1, 2), R106 = c(3, 4, 4, 4, 3, 3, 4, 4, 4, 4, 4, 
4, 3, 4, 4), R107 = c(1, 3, 3, 1, 2, 3, 2, 3, 3, 3, 3, 3, 1, 
3, 3), R108 = c(0, 4, 2, 2, 1, 3, 3, 2, 3, 3, 4, 3, 3, 3, 3), 
    R109 = c(0, 2, 0, 1, 1, 1, 2, 1, 1, 1, 1, 1, 2, 1, 1), R110 = c(1, 
    1, 0, 1, 1, 1, 2, 1, 0, 1, 0, 1, 0, 1, 1), R111 = c(3, 2, 
    2, 3, 3, 2, 2, 3, 1, 3, 4, 2, 2, 3, 2), R112 = c(1, 2, 2, 
    1, 1, 1, 1, 3, 1, 2, 2, 2, 1, 1, 1), Gender = structure(c(2L, 
    1L, 1L, 1L, 1L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = c("male", 
    "female"), class = "factor")), .Names = c("UID", "R101", 
"R102", "R103", "R104", "R105", "R106", "R107", "R108", "R109", 
"R110", "R111", "R112", "Gender"), row.names = c(NA, -15L), class = c("tbl_df", 
"tbl", "data.frame"))

p <-heatmaply(data4[1:13],fontsize_row = 8,fontsize_col = 6,Rowv =F,grid_gap = 0.5,colors = viridis(n = 256, alpha = 1, begin = 1,end = 0, option = "viridis"),branches_lwd = 0.2,row_side_colors =as.factor( data4$Gender))
p

btw - for discrete heatmaps, you can fake a discrete colour scale. See the code [here](https://github.com/talgalili/heatmaply/blob/master/R/plots.R#L636). Sorry for the lack of comments in the code, I could write a gist to explain the process in future. — alan ocallaghan, May 12 '18 at 12:50
Thanks. I data frame tip did the trick for me I removed the UID coloumn and the gender color is now discrete. — Santam, May 13 '18 at 17:15

score 0 · Accepted Answer · answered May 05 '18 at 15:02

The code does produce a colour-coded annotation for both factors. However, when there are enough levels, the colour scheme defaults to a rainbow scheme, which can be difficult to distinguish. You may need to try to subset the heatmap, or to try setting a different row_side_palette in heatmaply.

You might also want to pass the row_side_colors as a data.frame rather than a vector to ensure they are named correctly in both rownames and hovertext.

See the code below, which includes a few minor tweaks.

heatmaply(
  data4[, setdiff(colnames(data4), c("Gender", "UID"))],
  plot_method = "plotly",
  fontsize_row = 8,
  fontsize_col = 6,
  Rowv = FALSE,
  grid_gap = 0.5,
  colors = viridis(n = 256, alpha = 1, begin = 1,end = 0, option = "viridis"),
  branches_lwd = 0.2,
  row_side_colors = data4[, c("Gender", "UID")])

How to have an additional coloumn in a heatmaply heatmap

1 Answers1