R: geom_logo for multiple genes from tibble

user_12806025 0
Viewed 130

I am not familiar so much with ggseqlogo package so I would appreciate any form of help.

I have prepared tibble that looks like:

test <- tibble( gene = c("A", "B", "A", "C", "B"),
                seq = c("AAAAAAAAAAAAAAAAAAAA",
                        "GGGGGGGGGGGGGGGG",
                        "AAAAAATAAAAATAAAAAAA",
                        "AGTCGTCATGCATCAATCCCAATGGTGCA",
                        "GGGGGGGCCGGGGGGG") )

I want to prepare seqlogo per each gene, based on the gene name. Per each gene sequences have the same length.

As far I've tried this:

ggplot() + 
 geom_logo(data = test$gene) +
 facet_grid(rows = ~ gene)

But as far this is the best what I got:

tidyverse r sequence-diagram
1 Answers

The response might be late but works.

ggseqlogo has the option for facets but that requires unique gene names and equal size sequences. I avoided this issue by creating a loop and storing plots for all genes in a list.

That plot list can then be arranged in via cowplots plot_grid

library(tidyverse)
library(cowplot)
library(ggseqlogo)


test <- tibble( gene = c("A", "B", "A", "C", "B"),
                seq = c("AAAAAAAAAAAAAAAAAAAA",
                        "GGGGGGGGGGGGGGGG",
                        "AAAAAATAAAAATAAAAAAA",
                        "AGTCGTCATGCATCAATCCCAATGGTGCA",
                        "GGGGGGGCCGGGGGGG") ) 

# Initialize list to store plots in
plot_list <- list()

# Loop through all genes and 
# store the resulting plots in the plotlist
for(i in 1:nrow(test)) {
  plot_list[[i]] <- ggplot() + 
    geom_logo(data = test[i,2],  seq_type = "dna") +
    ggtitle(paste0( test[i,1]))
}

# Cowplots can arrange the list based on your desire
plot_grid(plotlist =  plot_list, ncol = 2)

enter image description here

user_129569470
Related