Supplementary file S02. Semantic similarity matrices of random GO terms

In this supplementary file, we demonstrate the heatmaps of semantic similarities for randomly sampled 500 GO terms, from Biological Process (BP), Molecular Function (MF) and Cellular Component (CC) ontologies. For each ontology category, we show 11 examples. The aim is to demonstrate similarity matrices of random GO terms are very common to show diagonal block patterns.

library(simplifyEnrichment)
library(grid)
library(circlize)
library(ComplexHeatmap)
library(GetoptLong)
library(cowplot)

set.seed(123)
col_fun = colorRamp2(c(0, 1), c("white", "red"))
lgd = Legend(title = "Similarity", col_fun = col_fun)

500 random BP

pl = list()
for(i in 1:11) {
    go_id = random_GO(500, "BP")
    mat = GO_similarity(go_id)
    pl[[i]] = grid.grabExpr(draw(Heatmap(mat, col = col_fun, show_row_names = FALSE, show_column_names = FALSE,
        show_row_dend = FALSE, show_column_dend = FALSE, show_heatmap_legend = FALSE, 
        column_title = qq("random BP #@{i}"))))
}
pl[[12]] = lgd@grob
plot_grid(plotlist = pl, nrow = 3)

500 random MF

pl = list()
for(i in 1:11) {
    go_id = random_GO(500, "MF")
    mat = GO_similarity(go_id)
    pl[[i]] = grid.grabExpr(draw(Heatmap(mat, col = col_fun, show_row_names = FALSE, show_column_names = FALSE,
        show_row_dend = FALSE, show_column_dend = FALSE, show_heatmap_legend = FALSE, 
        column_title = qq("random MF #@{i}"))))
}
pl[[12]] = lgd@grob
plot_grid(plotlist = pl, nrow = 3)

500 random CC

pl = list()
for(i in 1:11) {
    go_id = random_GO(500, "CC")
    mat = GO_similarity(go_id)
    pl[[i]] = grid.grabExpr(draw(Heatmap(mat, col = col_fun, show_row_names = FALSE, show_column_names = FALSE,
        show_row_dend = FALSE, show_column_dend = FALSE, show_heatmap_legend = FALSE, 
        column_title = qq("random CC #@{i}"))))
}
pl[[12]] = lgd@grob
plot_grid(plotlist = pl, nrow = 3)