Move data to a separate package

DESCRIPTION

@@ -1,37 +1,40 @@
 Type: Package
 Package: msigdbr
 Title: MSigDB Gene Sets for Multiple Organisms in a Tidy Data Format
-Version: 2023.1.1
+Version: 9.0.0.9000
 Authors@R: 
     person("Igor", "Dolgalev", , "[email protected]", role = c("aut", "cre"),
            comment = c(ORCID = "0000-0003-4451-126X"))
 Description: Provides the 'Molecular Signatures Database' (MSigDB) gene
     sets typically used with the 'Gene Set Enrichment Analysis' (GSEA)
     software (Subramanian et al. 2005 <doi:10.1073/pnas.0506580102>,
-    Liberzon et al. 2015 <doi:10.1016/j.cels.2015.12.004>) in a standard R
-    data frame with key-value pairs. The package includes the human genes
-    as listed in MSigDB as well as the corresponding symbols and IDs for
-    frequently studied model organisms such as mouse, rat, pig, fly, and
-    yeast.
+    Liberzon et al. 2015 <doi:10.1016/j.cels.2015.12.004>, Castanza et al.
+    2023 <doi:10.1038/s41592-023-02014-7>) as an R data frame. The package
+    includes the human genes as listed in MSigDB as well as the
+    corresponding symbols and IDs for frequently studied model organisms
+    such as mouse, rat, pig, fly, and yeast.
 License: MIT + file LICENSE
 URL: https://igordot.github.io/msigdbr/
 BugReports: https://github.com/igordot/msigdbr/issues
 Depends:
-    R (>= 3.6)
+    R (>= 4.1)
 Imports:
     babelgene (>= 22.9),
     dplyr (>= 1.1.1),
-    magrittr,
+    lifecycle,
+    methods,
     rlang,
     tibble,
-    tidyselect
+    tidyselect (>= 1.2.0)
 Suggests:
     knitr,
+    msigdbdf,
     rmarkdown,
     roxygen2,
     testthat
 VignetteBuilder: 
     knitr
+Additional_repositories: https://igordot.r-universe.dev
 Encoding: UTF-8
 Roxygen: list(markdown = TRUE)
 RoxygenNote: 7.3.2

NAMESPACE

@@ -1,11 +1,10 @@
 # Generated by roxygen2: do not edit by hand
 
-export("%>%")
 export(msigdbr)
 export(msigdbr_collections)
-export(msigdbr_show_species)
 export(msigdbr_species)
 import(tibble)
+import(tidyselect)
 importFrom(babelgene,orthologs)
 importFrom(babelgene,species)
 importFrom(dplyr,arrange)
@@ -16,8 +15,10 @@ importFrom(dplyr,inner_join)
 importFrom(dplyr,mutate)
 importFrom(dplyr,rename)
 importFrom(dplyr,select)
-importFrom(magrittr,"%>%")
+importFrom(lifecycle,deprecated)
+importFrom(lifecycle,is_present)
+importFrom(methods,is)
 importFrom(rlang,.data)
-importFrom(tibble,as_tibble)
-importFrom(tidyselect,any_of)
-importFrom(tidyselect,everything)
+importFrom(rlang,check_installed)
+importFrom(utils,install.packages)
+importFrom(utils,menu)

R/msigdbr-check-data.R

@@ -0,0 +1,50 @@
+#' Check that the data package is installed
+#'
+#' Check that the 'msigdbdf' data package is installed.
+#' If not, provide instructions for installation.
+#' A dependency listed in DESCRIPTION Suggests is not guaranteed to be installed.
+#'
+#' @importFrom utils install.packages menu
+msigdbr_check_data <- function() {
+  if (!requireNamespace("msigdbdf", quietly = TRUE)) {
+    message("The 'msigdbdf' package must be installed.")
+
+    install_instructions <- paste0(
+      "Please run the following command to install the 'msigdbdf' package:\n",
+      "install.packages('msigdbdf', repos = 'https://igordot.r-universe.dev')"
+    )
+
+    error_message <- function(e) {
+      message(e)
+      cat(paste0("\nFailed to install the 'msigdbdf' package.\n", install_instructions, "\n"))
+    }
+
+    if (interactive()) {
+      # If running R interactively
+      input <- utils::menu(c("Yes", "No"), title = "Would you like to install 'msigdbdf'?")
+      if (input == 1) {
+        # Answered "Yes"
+        message("Installing the 'msigdbdf' package.")
+        tryCatch(
+          utils::install.packages("msigdbdf", repos = c("https://igordot.r-universe.dev", getOption("repos"))),
+          error = error_message, warning = error_message
+        )
+      } else {
+        # Answered "No"
+        stop(install_instructions)
+      }
+    } else {
+      # If not running R interactively
+      stop(install_instructions)
+    }
+  }
+}
+
+.onAttach <- function(libname, pkgname) {
+  if (!requireNamespace("msigdbdf", quietly = TRUE)) {
+    packageStartupMessage(
+      "To access all the data, please install the 'msigdbdf' package with:\n",
+      "install.packages('msigdbdf', repos = 'https://igordot.r-universe.dev')"
+    )
+  }
+}

R/msigdbr-collections.R

@@ -2,14 +2,40 @@
 #'
 #' @return A data frame of the available collections.
 #'
+#' @param db_species Species abbreviation for the human or mouse databases (`"Hs"` or `"Mm"`).
+#'
 #' @importFrom dplyr arrange count distinct
 #' @export
 #'
 #' @examples
 #' msigdbr_collections()
-msigdbr_collections <- function() {
-  msigdbr_genesets %>%
-    distinct(.data$gs_cat, .data$gs_subcat, .data$gs_id) %>%
-    count(.data$gs_cat, .data$gs_subcat, name = "num_genesets") %>%
-    arrange(.data$gs_cat, .data$gs_subcat)
+msigdbr_collections <- function(db_species = "Hs") {
+  # rlang::check_installed("msigdbdf")
+  msigdbr_check_data()
+
+  # Get the full table of gene sets and their member genes
+  mc <- msigdbdf::msigdbdf(target_species = db_species)
+
+  # Keep only gene set information (ignors genes)
+  mc <- dplyr::distinct(
+    mc,
+    .data$gs_collection,
+    .data$gs_subcollection,
+    .data$gs_collection_name,
+    .data$gs_id
+  )
+
+  # Count the number of gene sets per collection
+  mc <- dplyr::count(
+    mc,
+    .data$gs_collection,
+    .data$gs_subcollection,
+    .data$gs_collection_name,
+    name = "num_genesets"
+  )
+
+  # Sort
+  mc <- dplyr::arrange(mc, .data$gs_collection, .data$gs_subcollection)
+
+  return(mc)
 }

R/msigdbr-package.R

@@ -0,0 +1,11 @@
+#' @keywords internal
+"_PACKAGE"
+
+## usethis namespace: start
+#' @import tibble
+#' @import tidyselect
+#' @importFrom lifecycle deprecated is_present
+#' @importFrom methods is
+#' @importFrom rlang .data check_installed
+## usethis namespace: end
+NULL

R/msigdbr-species.R

@@ -4,31 +4,19 @@
 #'
 #' @importFrom babelgene species
 #' @importFrom dplyr arrange distinct select
-#' @importFrom tibble as_tibble
+#'
 #' @export
 #'
 #' @examples
 #' msigdbr_species()
 msigdbr_species <- function() {
-  species() %>%
-    as_tibble() %>%
+  babelgene::species() |>
+    as_tibble() |>
     select(
       species_name = "scientific_name",
       species_common_name = "common_name"
-    ) %>%
-    rbind(c("Homo sapiens", "human")) %>%
-    distinct() %>%
+    ) |>
+    rbind(c("Homo sapiens", "human")) |>
+    distinct() |>
     arrange(.data$species_name)
 }
-
-#' List the species available in the msigdbr package
-#'
-#' This function is being deprecated and replaced by `msigdbr_species()`.
-#'
-#' @return A vector of possible species.
-#'
-#' @export
-msigdbr_show_species <- function() {
-  .Deprecated("msigdbr_species")
-  sort(msigdbr_species()[["species_name"]])
-}