Hi,

I have multiple timepoint samples pre- and post-treatment for several patients. I cannot see a standard way to colour mutation spectrum by a 'group' variable, as opposed to sample or mutation. I have the group in the VRanges object supplied to mutationContext, but cannot force via scale_fill_brewer or other such.

Is there a simple way to do this? I had a look in the ggplot2 object ('p' in the tutorial), this yields p$data, p$labels$fill. I could try and hack that to allow a group variable in p$data, but would be neater if there is a proposed solution from authors/others.

Thanks,

Bruce.

This is an interesting use case - it will not work out of the box, but we can make it happen.

I have created a modified version of the plotMutationSpectrum function that has the extra functionality:

library(SomaticSignatures)
library(ggplot2)
plotMutationSpectrumWithColorMapping <- function(vr, group,
    colorby = c("sample", "alteration", "colormapping"),
    normalize = TRUE, colormapping = NULL)
{
  colorby = match.arg(colorby)
  x = motifMatrix(vr, group, normalize = normalize)
  w_df = reshape2::melt(x, varnames = c("motif", "sample"))
  w_df$colormapping = colormapping[w_df$sample]
  w_df$alteration = sub("([ACGTN])([ACGTN]) .+", "\\1>\\2", w_df$motif)
  w_df$context = sub("[ACGTN][ACGTN] (.+)", "\\1", w_df$motif)
  p = ggplot(w_df)
  p = p + geom_bar(aes_string(x = "context", y = "value", fill = colorby),
                   stat = "identity", position = "identity")
  p = p + facet_grid(sample ~ alteration)
  p = p + SomaticSignatures:::theme_ss() + SomaticSignatures:::theme_small_axis()
  p = p + theme(legend.position = "none")
  p = p + scale_fill_brewer(palette = "Set3")
  p = p + xlab("Motif") + ylab("Contribution")
  return(p)
}

Let's try it with a small example data set:

library(SomaticSignatures)
library(ggplot2)
data(sca_motifs_tiny)

We want to split the data set according to cancer type and additionally distinguish between lung and non-lung cancer via color. For this, we define a mapping between our grouping (cancer studies) and our coloring (lung cancers) variable:

studies = as.character(unique(sca_motifs_tiny$study))
groups = ifelse(grepl("^L", colormapping), "lung", "non-lung")
colormapping = setNames(groups, studies)

      KIRC       SKCM       LUAD       LUSC       HNSC        GBM       THCA 
"non-lung" "non-lung"     "lung"     "lung" "non-lung" "non-lung" "non-lung" 
        OV 
"non-lung" 

Plotting and coloring the mutational spectrum according to sample/group and alteration type works as before:

plotMutationSpectrumWithColorMapping(sca_motifs_tiny, "study", "sample") + scb


plotMutationSpectrumWithColorMapping(sca_motifs_tiny, "study", "alteration") + scb

Most importantly, we can now use the color mapping defined above:

plotMutationSpectrumWithColorMapping(sca_motifs_tiny, "study", "colormap", colormapping = cm) + scb

This is not perfect, but let me know if this is what you were looking for and if turns out to be helpful.