Make deduplication and gap filling optional.

MicheleBortol · MicheleBortol · commit 88ae6805ec83 · 2022-11-18T14:27:56.000+01:00
diff --git a/example.config b/example.config
@@ -2,15 +2,18 @@
 params.input_path = "input"
 params.output_path = "output"
 
+// Workflow control parameters
+params.fill_gaps			= true		// Set to false to skip MindaGap Gap Filling
+params.deduplicate			= true		// Set to false to skip MindaGap transcript deduplication
+params.do_zip				= true		// Set to false to skip making ImageJ ROIs (faster)
+params.segmentation_tool	= "mesmer"	// Use Mesmer (mesmer) or Cellpose (cellpose) for segmentation
+
 // MindaGap transcript deduplication parameters
 params.tile_size		= 2144	// Tile size (distance between gridlines)
 params.window_size		= 30	// Window arround gridlines to search for duplicates
 params.max_freq			= 400  // Maximum transcript count to calculate X/Y shifts (better to discard very common genes)
 params.min_mode			= 10	// Minumum occurances of ~XYZ_shift to consider it valid
 
-// Use Mesmer (mesmer) or Cellpose (cellpose) for segmentation
-params.segmentation_tool = "mesmer"
-
 // Cellpose Segmentation Parameters
 params.model_name = "cyto"
 params.probability_threshold = 1
@@ -25,5 +28,3 @@ params.small_objects_threshold   = 15		// Minimum object size. Default(15)
 params.fill_holes_threshold      = 15		// Max Size for hole filling. Default(15)
 params.radius                    = 2		// Undocuented in Mesmer. Default(2)
 
-// Set to false to skip making ImageJ ROIs (faster)
-params.do_zip =	true
diff --git a/main.nf b/main.nf
@@ -28,25 +28,31 @@ workflow {
 	counts = sample_metadata.counts.toSortedList().flatten().view()
 
 	// Gap Filling with MindaGap
-	gap_filling(samples, dapi)
-
-	filled_images = gap_filling.out.gap_filled_image
-		.toSortedList(compare_file_names).flatten().view()
+    if (params.fill_gaps == true){
+	    gap_filling(samples, dapi)
+	    images = gap_filling.out.gap_filled_image
+		    .toSortedList(compare_file_names).flatten().view()
+    } else {
+        images = dapi
+    }
 
     // Deduplicate transcripts with MindaGap
-	deduplicating(samples, counts, params.tile_size, params.window_size, \
-        params.max_freq, params.min_mode)
-
-	clean_transcripts = deduplicating.out.deduplicated_transcripts
-		.toSortedList(compare_file_names).flatten().view()
+	if (params.deduplicate == true){
+        deduplicating(samples, counts, params.tile_size, params.window_size, \
+            params.max_freq, params.min_mode)
+	    transcripts = deduplicating.out.deduplicated_transcripts
+		    .toSortedList(compare_file_names).flatten().view()
+    } else {
+        transcripts = counts
+    }
 
 	// Cell Segmentation
 	if(params.segmentation_tool == "cellpose"){
 		segmentation = cellpose_segmentation(sample_metadata.sample, \
 			params.model_name, params.probability_threshold, \
-			params.cell_diameter, filled_images)
+			params.cell_diameter, images)
 	}else if (params.segmentation_tool == "mesmer") {
-		segmentation = mesmer_segmentation(sample_metadata.sample, filled_images, \
+		segmentation = mesmer_segmentation(sample_metadata.sample, images, \
 			params.maxima_threshold, params.maxima_smooth, params.interior_threshold, \
 			params.interior_smooth, params.small_objects_threshold, \
 			params.fill_holes_threshold, params.radius)
@@ -64,7 +70,7 @@ workflow {
 	}    
 
     // Single Cell Data Extraction
-	sc_data_extraction(samples, cell_masks, clean_transcripts)
+	sc_data_extraction(samples, cell_masks, transcripts)
     single_cell_data = sc_data_extraction.out.sc_data
                 .toSortedList(compare_file_names).flatten().view()
 }
