Version 1.0.5

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: fenr
 Title: Fast functional enrichment for interactive applications
-Version: 1.1.4
+Version: 1.1.5
 Authors@R: person(
     given = "Marek",
     family = "Gierlinski",
@@ -53,8 +53,7 @@ Suggests:
     testthat,
     knitr,
     rmarkdown,
-    topGO,
-    org.Hs.eg.db
+    topGO
 RoxygenNote: 7.2.3
 VignetteBuilder: knitr
 LazyData: false

NEWS.md

@@ -133,4 +133,12 @@
 
  - Reinstated Bioplanet access, this time with graceful fail when the website is down.
  - Minor code changes.
+
+## Version 1.0.5
+
+ - Bug fix: if feature id - term id mapping is not unique (which can happen), features are duplicated in counting; fixed by `dplyr::distinct()` on mapping
+ - Correction in vignette: using yeast genome for `topGO`, instead of human.
+ - Improving test coverage
+ - Making tests and examples resilient to Ensembl outage.
+
 

R/caching.R

@@ -84,10 +84,11 @@ cached_url_path <- function(rname, fpath, use_cache) {
 #' This function will remove all cached data used by `fenr`. The user will be
 #' prompted for confirmation. Use with caution!
 #'
+#' @param ask Logical, whether to ask user for confirmation.
 #' @return TRUE if successfully removed.
 #' @noRd
-remove_cache <- function() {
+remove_cache <- function(ask = TRUE) {
   cache <- cache_location()
   bfc <- BiocFileCache::BiocFileCache(cache, ask = FALSE)
-  BiocFileCache::removebfc(bfc, ask = TRUE)
+  BiocFileCache::removebfc(bfc, ask = ask)
 }

R/enrichment.R

@@ -107,7 +107,8 @@ prepare_for_enrichment <- function(terms, mapping, all_features = NULL,
   feature_term <- mapping |>
     dplyr::rename(feature_id = !!feature_name) |>
     dplyr::filter(feature_id %in% all_features) |>
-    dplyr::select(feature_id, term_id)
+    dplyr::select(feature_id, term_id) |>
+    dplyr::distinct()
 
   # Feature to terms hash
   f2t <- feature_term |>

R/go.R

@@ -322,8 +322,10 @@ fetch_go_from_bm <- function(mart, use_cache = TRUE) {
 #' @importFrom assertthat assert_that
 #' @examples
 #' # Fetch GO data from Ensembl
+#' \dontrun{
 #' mart <- biomaRt::useEnsembl(biomart = "ensembl", dataset = "scerevisiae_gene_ensembl")
 #' go_data_ensembl <- fetch_go(mart = mart, on_error = "warn")
+#' }
 #' # Fetch GO data from Gene Ontology
 #' go_data_go <- fetch_go(species = "sgd", on_error = "warn")
 fetch_go <- function(species = NULL, mart = NULL, use_cache = TRUE, on_error = c("stop", "warn")) {

tests/testthat/test_caching.R

@@ -62,3 +62,13 @@ test_that("Multiple entries behaviour", {
   expect_equal(re_content, expected_content)
 })
 
+
+test_that("Clearing cache", {
+  remove_cache(ask = FALSE)
+
+  cache <- cache_location()
+  bfc <- BiocFileCache::BiocFileCache(cache, ask = FALSE)
+  cnt <- BiocFileCache::bfccount(bfc)
+
+  expect_equal(cnt, 0)
+})

tests/testthat/test_go.R

@@ -98,9 +98,19 @@ test_that("GO yeast from Ensembl is correct", {
     "GO:0004365", "YGR192C"
   )
 
-  mart <- biomaRt::useEnsembl(biomart = "ensembl", dataset = "scerevisiae_gene_ensembl")
-  re <- fetch_go(mart = mart)
-  test_fetched_structure(re)
-  test_terms(re$terms, expected_terms)
-  test_mapping(re$mapping, expected_mapping, "ensembl_gene_id")
+  mart <- tryCatch(
+    {biomaRt::useEnsembl(biomart = "ensembl", dataset = "scerevisiae_gene_ensembl")},
+    error = function(cond) {
+      message("Biomart is not responding.")
+      message(conditionMessage(cond))
+      NULL
+    }
+  )
+
+  if(!is.null(mart)) {
+    re <- fetch_go(mart = mart)
+    test_fetched_structure(re)
+    test_terms(re$terms, expected_terms)
+    test_mapping(re$mapping, expected_mapping, "ensembl_gene_id")
+  }
 })

tests/testthat/test_prepare_for_enrichment.R

@@ -83,6 +83,21 @@ test_that("Expected correct output", {
 })
 
 
+test_that("Duplicate term-gene pairs are removed", {
+  extra_row <- mapping[1, ]
+  mapping_dup <- mapping |>
+    dplyr::add_row(extra_row)
+  td <- prepare_for_enrichment(terms, mapping_dup, feature_name = "feature_id")
+  returned <- td$term2feature[[extra_row$term_id]] |>
+    sort()
+  expected <- mapping |>
+    dplyr::filter(term_id == extra_row$term_id) |>
+    dplyr::pull(feature_id) |>
+    sort()
+  expect_equal(expected, returned)
+})
+
+
 test_that("Wrong columns in terms", {
   trms <- terms |>
     dplyr::rename(gobble = term_id)

vignettes/fenr.Rmd

@@ -159,7 +159,6 @@ In exploratory data analysis, one often conducts several enrichment analyses. Th
 ```{r topGO}
 suppressPackageStartupMessages({
   library(topGO)
-  library(org.Hs.eg.db)
 })
 
 # Preparing the gene set
@@ -171,9 +170,14 @@ gene_selection <- function(genes) {
   genes > 0
 }
 
-# Mapping genes to GO
-allGO2genes <- annFUN.org(whichOnto = "BP", feasibleGenes = NULL,
-                          mapping = "org.Hs.eg.db", ID = "symbol")
+# Mapping genes to GO, we use the go2gene downloaded above and convert it to a
+# list
+go2gene <- go$mapping |> 
+  dplyr::select(gene_symbol, term_id) |> 
+  dplyr::distinct() |> 
+  dplyr::group_by(term_id) |> 
+  dplyr::summarise(ids = list(gene_symbol)) |> 
+  tibble::deframe()
 
 # Setting up the GO data for analysis
 go_data <- new(
@@ -182,7 +186,7 @@ go_data <- new(
   allGenes = all_genes,
   geneSel = gene_selection,
   annot = annFUN.GO2genes,
-  GO2genes = allGO2genes
+  GO2genes = go2gene
 )
 
 # Performing the enrichment test