[r] get_seurat: allow selection of Seurat counts or data slot (#607)

Closes #596

api/r/cellxgene.census/R/get_helpers.R

@@ -22,7 +22,9 @@ get_presence_matrix <- function(census, organism, measurement_name = "RNA") {
 #' @param census The census object, usually returned by `cellxgene.census::open_soma()`.
 #' @param organism The organism to query, usually one of `Homo sapiens` or `Mus musculus`
 #' @param measurement_name The measurement object to query. Defaults to `RNA`.
-#' @param X_name The `X` layer to query. Defaults to `raw`.
+#' @param X_layers A named character of `X` layers to add to the Seurat assay, where the names are
+#'        the names of Seurat slots (`counts` or `data`) and the values are the names of layers
+#'        within `X`.
 #' @param obs_value_filter A SOMA `value_filter` across columns in the `obs` dataframe, expressed as string.
 #' @param obs_coords A set of coordinates on the obs dataframe index, expressed in any type or format supported by SOMADataFrame's read() method.
 #' @param obs_column_names Columns to fetch for the `obs` data frame.
@@ -39,7 +41,7 @@ get_seurat <- function(
     census,
     organism,
     measurement_name = "RNA",
-    X_name = "raw",
+    X_layers = c(counts = "raw", data = NULL),
     obs_value_filter = NULL,
     obs_coords = NULL,
     obs_column_names = NULL,
@@ -53,8 +55,7 @@ get_seurat <- function(
     var_query = tiledbsoma::SOMAAxisQuery$new(value_filter = var_value_filter, coords = var_coords)
   )
   return(expt_query$to_seurat(
-    # TODO: should we allow selection of the seurat 'counts' or 'data' slot?
-    c(counts = X_name),
+    X_layers = X_layers,
     obs_column_names = obs_column_names,
     var_column_names = var_column_names
   ))

api/r/cellxgene.census/tests/testthat/test-get_helpers.R

@@ -103,3 +103,33 @@ test_that("get_seurat allows missing obs or var filter", {
   expect_equal(ncol(seurat[["RNA"]]), 10001)
   expect_equal(nrow(seurat[["RNA"]]), 1)
 })
+
+test_that("get_seurat X_layers allows slot selection", {
+  census <- open_soma_latest_for_test()
+  on.exit(census$close(), add = TRUE)
+
+  # default: raw X layer gets loaded into counts (which Seurat implicitly copies into data)
+  seurat <- get_seurat(
+    census,
+    "Mus musculus",
+    obs_value_filter = "tissue_general == 'vasculature'",
+    obs_column_names = c("soma_joinid", "cell_type", "tissue", "tissue_general", "assay"),
+    var_value_filter = "feature_name %in% c('Gm53058', '0610010K14Rik')",
+    var_column_names = c("soma_joinid", "feature_id", "feature_name", "feature_length")
+  )
+  expect_equal(nrow(seurat$RNA@counts), 2)
+  expect_equal(nrow(seurat$RNA@data), 2)
+
+  # expressly load raw X data into data; then counts is empty
+  seurat <- get_seurat(
+    census,
+    "Mus musculus",
+    obs_value_filter = "tissue_general == 'vasculature'",
+    obs_column_names = c("soma_joinid", "cell_type", "tissue", "tissue_general", "assay"),
+    var_value_filter = "feature_name %in% c('Gm53058', '0610010K14Rik')",
+    var_column_names = c("soma_joinid", "feature_id", "feature_name", "feature_length"),
+    X_layers = c(counts = NULL, data = "raw")
+  )
+  expect_equal(nrow(seurat$RNA@counts), 0)
+  expect_equal(nrow(seurat$RNA@data), 2)
+})