From 4cf30cb49f58ce6dc05a398f488fb7e4b5ddc20c Mon Sep 17 00:00:00 2001
From: Ally Hawkins <ally.hawkins@ccdatalab.org>
Date: Wed, 15 Jan 2025 12:35:06 -0600
Subject: [PATCH] account for entire projects with cell lines

---
 .../usr/bin/combine-celltype-tables.R         | 50 +++++++++++--------
 1 file changed, 28 insertions(+), 22 deletions(-)

diff --git a/modules/cell-type-consensus/resources/usr/bin/combine-celltype-tables.R b/modules/cell-type-consensus/resources/usr/bin/combine-celltype-tables.R
index f562fd6..dc79f5f 100755
--- a/modules/cell-type-consensus/resources/usr/bin/combine-celltype-tables.R
+++ b/modules/cell-type-consensus/resources/usr/bin/combine-celltype-tables.R
@@ -91,25 +91,31 @@ blueprint_df <- data.frame(
 
 # Create combined TSV ----------------------------------------------------------
 
-# read in TSV files and combine into a single df
-all_cells_df <- all_files |>
-  purrr::map(readr::read_tsv) |>
-  dplyr::bind_rows() |>
-  # add columns for panglao ontology and consensus
-  # first add panglao ontology
-  dplyr::left_join(panglao_ref_df, by = c("cellassign_celltype_annotation" = "original_panglao_name")) |>
-  # now add in all the blueprint columns
-  dplyr::left_join(blueprint_df, by = c("singler_celltype_ontology" = "blueprint_ontology")) |>
-  # then add consensus labels
-  dplyr::left_join(consensus_ref_df,
-    by = c(
-      "singler_celltype_ontology" = "blueprint_ontology",
-      "cellassign_celltype_annotation" = "original_panglao_name",
-      "panglao_ontology"
-    )
-  ) |>
-  # use unknown for NA annotation but keep ontology ID as NA
-  dplyr::mutate(consensus_annotation = dplyr::if_else(is.na(consensus_annotation), "Unknown", consensus_annotation))
-
-# export file
-readr::write_tsv(all_cells_df, opt$output_file)
+# account for all samples being cell lines and no cell type annotations being present
+if (length(all_files) == 0) {
+  # make an empty filtered file
+  file.create(opt$output_file)
+} else {
+  # read in TSV files and combine into a single df
+  all_cells_df <- all_files |>
+    purrr::map(readr::read_tsv) |>
+    dplyr::bind_rows() |>
+    # add columns for panglao ontology and consensus
+    # first add panglao ontology
+    dplyr::left_join(panglao_ref_df, by = c("cellassign_celltype_annotation" = "original_panglao_name")) |>
+    # now add in all the blueprint columns
+    dplyr::left_join(blueprint_df, by = c("singler_celltype_ontology" = "blueprint_ontology")) |>
+    # then add consensus labels
+    dplyr::left_join(consensus_ref_df,
+      by = c(
+        "singler_celltype_ontology" = "blueprint_ontology",
+        "cellassign_celltype_annotation" = "original_panglao_name",
+        "panglao_ontology"
+      )
+    ) |>
+    # use unknown for NA annotation but keep ontology ID as NA
+    dplyr::mutate(consensus_annotation = dplyr::if_else(is.na(consensus_annotation), "Unknown", consensus_annotation))
+
+  # export file
+  readr::write_tsv(all_cells_df, opt$output_file)
+}