Approaches to add tracks / layers / grids to circlize::chordDiagram with specifications

[chrmirian]

Hi Zuguang Gu (and others)

Is it possible to add several tracks / layers to a chordDiagram?

I need to visualize additional information on my chordDiagram. I tried highlight.section, but that did not seem to work as straightforward for my requirements.

Let's use this data.

set.seed(1)

  df <- data.frame(
    Genes = 1:52,
    Description = sample(LETTERS[1:7], size = 52, replace = TRUE),
    value1 = sample(0:1, size = 52, replace = TRUE),
    value2 = sample(runif(52, min = 0, max = 0.001), size = 52, replace = TRUE),
    value3 = sample(10:20, size = 52, replace = TRUE)
  )

I have made this plot using the code below

matrix <- with(df, table(Description, Genes))  
  
  
  circos.clear()
  circos.par(start.degree = 90)
  
  description_colors <- setNames(viridis(length(rownames(matrix))), rownames(matrix))
  col_names_color <- setNames(rep("grey", length(colnames(matrix))), colnames(matrix))
  all_colors <- c(description_colors, col_names_color)
  
  # Generate the chord diagram with specified colors
  chordDiagram(matrix, transparency = 0.5, 
               annotationTrack = "grid", 
               annotationTrackHeight = c(0.03),
               preAllocateTracks = list(track.height = 0.1), # Reduced track height for genes
               grid.col = all_colors, # Apply colors to both descriptions and genes
               directional = -1,
               big.gap = 30, small.gap = 1)  # Adjust the highlight sector height here
  

  
  # Text labels for the sectors
  circos.trackPlotRegion(track.index = 1, panel.fun = function(x, y) {
    xlim = get.cell.meta.data("xlim")
    ylim = get.cell.meta.data("ylim")
    sector.name = get.cell.meta.data("sector.index")
    circos.text(CELL_META$xcenter, ylim[1] + cm_h(2), sector.name, 
                facing = "clockwise", niceFacing = TRUE, adj = c(0, 0.5))
  }, bg.border = NA)

Question: As you can see, df also have $value1, $value2, $value3. I want to add something like this:

The df$value could denote log2-fold change or FDR. Therefore, a gradient color legend should be added and correspond to each of the 3 values.

OBS: The df$value3 has been drawn "below" / on the inside of the "directional" color. It may also be drawn on the "outside" of the grey colored areas of each df$Gene so the directional color maintain a clear connection to each Description.

I also posted this on SO, but figured this site might be more appropriate.