Merge pull request #107 from AlexsLemonade/jashapiro/slim-sim

modules/simulate-sce/readme.md

@@ -4,4 +4,4 @@ This workflow is designed to simulate single cell data, primarily using the [spl
 
 Scripts are derived from the the `simulate-sce` module of the [OpenScPCA-analysis](https://github.com/AlexsLemonade/OpenScPCA-analysis) repository.
 
-Permalink to the version used: https://github.com/AlexsLemonade/OpenScPCA-analysis/tree/714282f90979400528a213b11ea71718bdd3527c/analyses/simulate-sce
+Permalink to the version used: https://github.com/AlexsLemonade/OpenScPCA-analysis/tree/a3d8a2c9144e8edb3894a7beeb89cdc6c3e6d681/analyses/simulate-sce

modules/simulate-sce/resources/usr/bin/simulate-sce.R

@@ -134,11 +134,16 @@ simulate_sce <- function(sce, ncells, replacement_metadata, processed) {
   # reduce the cell type data matrices, if present
   if (!is.null(metadata(sce_sim)$singler_results)) {
     sim_cells <- cell_subset[cell_subset %in% rownames(metadata(sce_sim)$singler_results)]
-    metadata(sce_sim)$singler_results <- metadata(sce_sim)$singler_results[sim_cells, ]
+    singler_results <- metadata(sce_sim)$singler_results[sim_cells, ]
+    # remove any gene data present to save space
+    metadata(singler_results)$de.genes <- NULL
+    metadata(singler_results)$common.genes <- NULL
+    metadata(sce_sim)$singler_results <- singler_results
   }
   if (!is.null(metadata(sce_sim)$cellassign_predictions)) {
-    sim_cells <- cell_subset[cell_subset %in% rownames(metadata(sce_sim)$cellassign_predictions)]
-    metadata(sce_sim)$cellassign_predictions <- metadata(sce_sim)$cellassign_predictions[sim_cells, ]
+    cellassign_subset <- metadata(sce_sim)$cellassign_predictions |>
+      dplyr::filter(barcode %in% cell_subset)
+    metadata(sce_sim)$cellassign_predictions <- cellassign_subset
   }
 
   # Adjust cluster/cell type labels --------------------------------------------