Code cleanup

assets/schema_input.json

@@ -21,14 +21,16 @@
             "spot_table": {
                 "type": "string",
                 "pattern": "^\\S+\\.(txt|tsv)$",
+                "format": "file-path",
                 "errorMessage": "Spot table must be provided, has to have shape x,y,z,gene with sep = '\t', cannot contain spaces and must have extension '.txt'"
             },
             "membrane_image": {
                 "errorMessage": "Membrane image is optional, and cannot contain spaces and must have extension '.tif' or '.tiff'",
                 "anyOf": [
                     {
                         "type": "string",
-                        "pattern": "^\\S+\\.(tif|tiff)$"
+                        "pattern": "^\\S+\\.(tif|tiff)$",
+                        "format": "file-path"
                     },
                     {
                         "type": "string",

modules/local/createanndata.nf

@@ -17,13 +17,12 @@ process CREATE_ANNDATA {
     script:
     def args   = task.ext.args   ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
-
     """
     create_anndata.py \\
         --input ${spot2cell} \\
         --spatial_cols X_centroid Y_centroid \\
         --output ${prefix}.adata \\
-        $args
+        ${args}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

modules/local/createstack.nf

@@ -17,12 +17,11 @@ process CREATE_STACK {
     script:
     def args   = task.ext.args   ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
-
     """
     stack.py \\
         --input ${image} \\
         --output ${prefix}.ome.tif \\
-        $args
+        ${args}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

modules/local/crophdf5.nf

@@ -16,15 +16,13 @@ process CROPHDF5 {
     task.ext.when == null || task.ext.when
 
     script:
-    def args   = task.ext.args   ?: ''
-    def prefix = task.ext.prefix ?: "${meta.id}"
-
+    def args = task.ext.args ?: ''
     """
     crop_hdf5.py \\
-        --input $image_stack \\
+        --input ${image_stack} \\
         --output . \\
-        --num_channels $num_channels \\
-        $args
+        --num_channels ${num_channels} \\
+        ${args}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

modules/local/croptiff.nf

@@ -5,8 +5,7 @@ process CROPTIFF {
     container 'ghcr.io/schapirolabor/molkart-local:v0.0.4'
 
     input:
-    tuple val(meta), path(image_stack)
-    tuple val(meta), path(crop_summary)
+    tuple val(meta), path(image_stack), path(crop_summary)
 
     output:
     tuple val(meta), path("*.tiff"), emit: crop_tiff
@@ -17,14 +16,12 @@ process CROPTIFF {
     task.ext.when == null || task.ext.when
 
     script:
-    def args   = task.ext.args   ?: ''
-    def prefix = task.ext.prefix ?: "${meta.id}"
-
+    def args = task.ext.args ?: ''
     """
     crop_tiff.py \\
-        --input $image_stack \\
-        --crop_summary $crop_summary \\
-        $args
+        --input ${image_stack} \\
+        --crop_summary ${crop_summary} \\
+        ${args}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

modules/local/maskfilter.nf

@@ -33,7 +33,6 @@ process MASKFILTER {
     """
 
     stub:
-    def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
 
     """

modules/local/molkartqc.nf

@@ -20,13 +20,13 @@ process MOLKARTQC{
 
     """
     collect_QC.py \\
-        --cellxgene $cellxgene_table \\
-        --spots $spot_table \\
-        --sample_id $prefix \\
-        --segmentation_method $segmethod \\
-        --filterqc $filterqc \\
+        --cellxgene ${cellxgene_table} \\
+        --spots ${spot_table} \\
+        --sample_id ${prefix} \\
+        --segmentation_method ${segmethod} \\
+        --filterqc ${filterqc} \\
         --outdir . \\
-        $args
+        ${args}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
@@ -35,7 +35,6 @@ process MOLKARTQC{
     """
 
     stub:
-    def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
 
     """

modules/local/molkartqcpng.nf

@@ -18,9 +18,9 @@ process MOLKARTQCPNG {
 
     """
     collect_QC.py \\
-        --png_overview $png \\
+        --png_overview ${png} \\
         --outdir . \\
-        $args
+        ${args}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
@@ -29,9 +29,7 @@ process MOLKARTQCPNG {
     """
 
     stub:
-    def args = task.ext.args ?: ''
-    def prefix = task.ext.prefix ?: "${meta.id}"
-
+    def prefix = task.ext.prefix ?: ( png.name.toString().tokenize('.')[0] )
     """
     touch ${prefix}.png
 

modules/local/spot2cell.nf

@@ -6,8 +6,7 @@ process SPOT2CELL{
     container 'ghcr.io/schapirolabor/molkart-local:v0.0.4'
 
     input:
-    tuple val(meta) , path(spot_table)
-    tuple val(meta2), path(cell_mask)
+    tuple val(meta), path(spot_table), path(cell_mask)
 
     output:
     tuple val(meta), path("*.csv"), emit: cellxgene_table

modules/local/tiffh5convert.nf

@@ -15,15 +15,13 @@ process TIFFH5CONVERT {
     task.ext.when == null || task.ext.when
 
     script:
-    def args   = task.ext.args   ?: ''
-    def prefix = task.ext.prefix ?: "${meta.id}"
-
+    def args = task.ext.args ?: ''
     """
     crop_hdf5.py \\
-        --input $image \\
+        --input ${image} \\
         --output . \\
-        --num_channels $num_channels \\
-        $args
+        --num_channels ${num_channels} \\
+        ${args}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

workflows/molkart.nf

@@ -44,31 +44,21 @@ workflow MOLKART {
 
     // stain: "1" denotes membrane, stain: "0" denotes nuclear image
     // this is used to preserve the order later
-    ch_samplesheet
-        .map {
-            meta, nuclear, spots, membrane ->
-            def new_meta = meta.clone()
-            new_meta.stain = "1"
-            return membrane != [] ? [new_meta, membrane] : null
-            //it[3] != [] ? tuple([id:it[0],stain:"1"], it[3]) : null
-        }.set { membrane_tuple } // if a membrane image is provided, return membrane image channel tuple (meta, path)
+    membrane_tuple = ch_samplesheet
+        .filter { it[3] } // filter samples with membrane
+        .map {meta, _nuclear, _spots, membrane ->
+            [meta + [stain: '1'], membrane]
+        }
 
-    ch_samplesheet
-        .map {
-            meta, nuclear, spots, membrane ->
-            def new_meta = meta.clone()
-            new_meta.stain = "0"
-            return [new_meta, nuclear]
+    image_tuple = ch_samplesheet
+        .map {meta, nuclear, _spots, _membrane ->
+            [meta + [stain: '0'], nuclear]
         }
-        .set { image_tuple } // creates nuclear image channel tuple (meta, path)
 
-    ch_samplesheet
-        .map {
-            meta, nuclear, spots, membrane ->
-            def new_meta = meta.clone()
-            return [meta, spots]
+    spot_tuple = ch_samplesheet
+        .map {meta, _nuclear, spots, _membrane ->
+            [meta, spots]
         }
-        .set { spot_tuple } // creates spot table channel tuple (meta, path)
 
     //
     // MODULE: Run Mindagap_mindagap
@@ -88,27 +78,25 @@ workflow MOLKART {
     map_for_stacks = params.skip_clahe ? clahe_in : CLAHE.out.img_clahe
 
     map_for_stacks
-        .map {
-            meta, tiff -> [meta.subMap("id"), tiff, meta.stain] // creates a channel containing only the sample id in meta, path to preprocessed image and the stain value ("0" or "1")
-        }.groupTuple() // combines based on meta
+        .map { meta, tiff ->
+            [ meta.subMap("id"), tiff, meta.stain ]
+        }
+        .groupTuple(by: 0)
+        .map { id, tiffs, stains ->
+            def sorted = [tiffs, stains].transpose().sort { it[1] }
+            def nuclear = sorted[0]
+            def membrane = sorted.size() > 1 ? sorted[1] : null
+            membrane ? [id, nuclear[0], membrane[0]] : [id, nuclear[0]]
+        }
+        .set{ grouped_map_stack }
+
+    grouped_map_stack.filter { !it[2] } // for rows without a present membrane image, set channel to no_stack
+        .set{ no_stack }
+
+    grouped_map_stack.filter{ it[2] }      // for rows where the membrane image is present, create a list of images to be stacked
         .map{
-            meta, paths, stains -> [meta, [paths[0], stains[0]], [paths[1], stains[1]]] // reorganizes to match path and stain
-        }.map{
-            meta, stain1, stain2 -> [meta, [stain1, stain2].sort{ it[1] }] // sort by stain index (0 for nuclear, 1 for other)
-        }.map{
-            meta, list -> [meta, list[0], list[1]] // sorted will have null as first list
-        }.map{
-            it[1][0] != null ? [it[0],it[1][0],it[2][0]] : [it[0],it[2][0]] // if null, only return the valid nuclear path value, otherwise return both nuclear and membrane paths
-        }.set { grouped_map_stack }
-
-    grouped_map_stack.filter{ // for rows without a present membrane image, set channel to no_stack
-        it[2] == null
-        }.set{ no_stack }
-
-    grouped_map_stack.filter{ // for rows where the membrane image is present, make it compliant with STACK inputs
-        it[2] != null
-        }.map{
-            [it[0],tuple(it[1],it[2])]
+            id, nuclear, membrane ->
+            [id, tuple(nuclear, membrane)]
         }.set{ create_stack_in }
 
     //
@@ -130,10 +118,7 @@ workflow MOLKART {
         ch_versions = ch_versions.mix(CROPHDF5.out.versions)
         // Combine images with crop_summary for making the same training tiff stacks as ilastik
         tiff_crop = stack_mix.join(CROPHDF5.out.crop_summary)
-        CROPTIFF(
-            tiff_crop.map(it -> tuple(it[0],it[1])),
-            tiff_crop.map(it -> tuple(it[0],it[2])),
-            )
+        CROPTIFF(tiff_crop)
         ch_versions = ch_versions.mix(CROPTIFF.out.versions)
         MOLKARTQCPNG(CROPTIFF.out.overview.map{
                     tuple('matchkey', it[1])
@@ -222,7 +207,7 @@ workflow MOLKART {
     }
     segmentation_masks.map{
         meta, mask, segmentation ->
-        new_meta = meta.clone()
+        def new_meta = meta.clone()
         new_meta.segmentation = segmentation
         [new_meta, mask]
     }.set { matched_segmasks }
@@ -244,16 +229,12 @@ workflow MOLKART {
         .combine(filtered_masks, by: 0)
         .map {
             meta, spots_table, mask, segmethod ->
-            new_meta = meta.clone()
+            def new_meta = meta.clone()
             new_meta.segmentation = segmethod
             [new_meta, spots_table, mask]
             }
         .set { dedup_spots }
-
-    SPOT2CELL(
-        dedup_spots.map(it -> tuple(it[0],it[1])),
-        dedup_spots.map(it -> tuple(it[0],it[2]))
-    )
+    SPOT2CELL(dedup_spots)
     ch_versions = ch_versions.mix(SPOT2CELL.out.versions)
 
     //