Update the internal dataset

R/msigdbr-check-data.R

@@ -7,7 +7,7 @@
 #' @importFrom utils install.packages menu
 msigdbr_check_data <- function() {
   if (!requireNamespace("msigdbdf", quietly = TRUE)) {
-    message("The 'msigdbdf' package must be installed to access all the data.")
+    message("The 'msigdbdf' package must be installed to access the full dataset.")
 
     install_instructions <- paste0(
       "Please run the following command to install the 'msigdbdf' package:\n",

R/msigdbr.R

@@ -60,12 +60,10 @@ msigdbr <- function(species = "Homo sapiens", db_species = "HS", collection = NU
     # Get the gene sets table
     mdb <- msigdbdf::msigdbdf(target_species = db_species)
   } else {
-    if (species == "Homo sapiens" && db_species == "HS" && collection == "H") {
-      # Use an internal human Hallmark dataset for minimal functionality without msigdbdf
-      # msigdb_h <- msigdbdf::msigdbdf(target_species = "HS")
-      # msigdb_h <- dplyr::filter(msigdb_h, gs_collection == "H")
-      # usethis::use_data(msigdb_h, internal = TRUE, overwrite = TRUE, compress = "xz")
-      mdb <- msigdb_h
+    if (species == "Homo sapiens" && db_species == "HS") {
+      # Use an internal dataset for minimal functionality without msigdbdf
+      mdb <- testdb
+      message("The 'msigdbdf' package must be installed to access the full dataset.")
     } else {
       # Check if msigdbdf is available and try to install otherwise
       msigdbr_check_data()

data-raw/test-dataset.R

@@ -0,0 +1,31 @@
+# Prepare the test dataset
+
+library(dplyr)
+
+# Get the full human dataset
+mdb <- msigdbdf::msigdbdf(target_species = "HS")
+
+# Get the Hallmark gene sets
+hallmark_gs_ids <- filter(mdb, gs_collection == "H") |> pull(gs_id)
+
+# Subsample smaller gene sets from every collection and sub-collection
+set.seed(99)
+random_gs_ids <- mdb |>
+  count(gs_id, gs_collection, gs_subcollection) |>
+  filter(n < 100) |>
+  group_by(gs_collection, gs_subcollection) |>
+  slice_sample(n = 5) |>
+  pull(gs_id)
+
+# Subset the full table to the
+subset_gs_ids <- unique(sort(c(hallmark_gs_ids, random_gs_ids)))
+testdb <- filter(mdb, gs_id %in% subset_gs_ids)
+
+# count(testdb, gs_collection, gs_subcollection)
+# count(testdb, db_gene_symbol, sort = TRUE)
+
+# Modify the version to indicate that this is a test dataset
+testdb$db_version <- paste0("TEST.", testdb$db_version)
+
+# Save package data
+usethis::use_data(testdb, internal = TRUE, overwrite = TRUE, compress = "xz")