How to assign colours to specific genes in gggenes arrowplot?

Question

I am new to R and I am trying to make an arrowplot. However, the basic gggenes set3 colour theme only has 12 colours and I need more.

I want to assign a group of genes with colour (eg, glycosyltransferases all red and methyltransferases all blue)

I have added an extra column to my df named "colour" and assigned each gene with one hex code (#c1ffc1) - just to test that all genes could change colour before going through and assigning the ones for glycosyltransferases etc - I managed to get it to change colour once and now it isn't working?

Here is the code example with three genes

#add colour column to assign to genes
> colour <- c("#c1ffc1")
> df1$colour <- colour
> #change colour
> library(ggplot2)
> library(gggenes)
> ggplot(df1, aes(xmin = start, xmax = end, y = molecule, fill = colour)) +
+   geom_gene_arrow() +
+   geom_gene_label(aes(label = gene)) + 
+   facet_wrap(~ molecule, scales = "free", ncol = 1) + 
+   theme(legend.position="top") + xlim(0,37841) + scale_fill_discrete(name = "gene", labels = c("VanH", "VanA", "VanX"))

 molecule start   end  strand   gene  orientation  colour
 KJ364518.1  2314  3345 reverse vanH 1  #f15854
 KJ364518.1  3347  4387 reverse vanA 1  #f15854
 KJ364518.1  4384  4992 reverse vanX 1  #f15854
 KJ364518.1  6334  7125 reverse ajrR 1  #faa43a
 KJ364518.1  7246  8097 reverse pdh  1  #5da5da
 KJ364518.1  8410 10272 reverse tri  1  #b276b2

Thanks so much in advance, Lucy

Answer 1

Your code appears to work fine, the only thing is that you're providing literal colours to the mapping instead of fill = your_classification_here. To show literal colours you can use scale_fill_identity().

library(ggplot2)
library(gggenes)

txt <- " molecule start   end  strand   gene  orientation  colour
 KJ364518.1  2314  3345 reverse vanH 1  #f15854
 KJ364518.1  3347  4387 reverse vanA 1  #f15854
 KJ364518.1  4384  4992 reverse vanX 1  #f15854
 KJ364518.1  6334  7125 reverse ajrR 1  #faa43a
 KJ364518.1  7246  8097 reverse pdh  1  #5da5da
 KJ364518.1  8410 10272 reverse tri  1  #b276b2"

df1 <- data.table::fread(text = txt)

ggplot(df1, aes(xmin = start, xmax = end, y = molecule, fill = colour)) +
  geom_gene_arrow() +
  geom_gene_label(aes(label = gene)) +
  facet_wrap(~ molecule, scales = "free", ncol = 1) +
  scale_fill_identity()

^{Created on 2022-05-11 by the reprex package (v2.0.1)}