Changes to spatialdata handling based on coordinate space

modules/local/ficture/model/main.nf

@@ -19,13 +19,13 @@ process FICTURE {
     script:
     def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
-    def features = features ? "--in-feature ${features}": ""
+    def features_list = features ? "--in-feature ${features}": ""
 
     """
     ficture run_together \\
         --in-tsv ${transcripts} \\
         --in-minmax ${coordinate_minmax} \\
-        ${features} \\
+        ${features_list} \\
         --out-dir results \\
         --train-width 12,18 \\
         --n-factor 6,12 \\

modules/local/spatialdata/write/main.nf

@@ -8,6 +8,7 @@ process SPATIALDATA_WRITE {
     tuple val(meta), path(bundle, stageAs: "*")
     val(outputfolder)
     val(segmented_object)
+    val(coordinate_space)
 
     output:
     tuple val(meta), path("${outputfolder}"), emit: spatialdata

modules/local/spatialdata/write/templates/write.py

@@ -13,8 +13,9 @@ def main():
     output_path = "."
     outputfolder = "${outputfolder}"
     segmented_object = "${segmented_object}"
+    coordinate_space = "${coordinate_space}"
 
-    cells_as_circles=True
+    cells_as_circles=False
     cells_boundaries=False
     nucleus_boundaries=False
     cells_labels=False
@@ -32,7 +33,19 @@ def main():
         cells_labels=True
         nucleus_labels=True
     else:
-        cells_as_circles=False
+        cells_as_circles = False
+
+    # set sd variables based on the coordinate space
+    if ( coordinate_space == "pixels" ):
+        cells_labels = True
+        nucleus_labels = True
+
+    if ( coordinate_space == "microns" ):
+        cells_labels = False
+        cells_boundaries = True
+        nucleus_boundaries = False
+        nucleus_labels = False
+        cells_as_circles = False
 
     format = "${params.format}"
     if ( format == "xenium" ):

nextflow.config

@@ -270,7 +270,6 @@ dag {
 
 manifest {
     name            = 'nf-core/spatialxe'
-    author          = """Florian Heyl""" // The author field is deprecated from Nextflow version 24.10.0, use contributors instead
     contributors    = [
         [
             name: 'Sameesh Kher',

subworkflows/local/baysor_run_prior_segmentation_mask/main.nf

@@ -27,6 +27,7 @@ workflow BAYSOR_RUN_PRIOR_SEGMENTATION_MASK {
     ch_htmls                = Channel.empty()
 
     ch_redefined_bundle     = Channel.empty()
+    ch_coordinate_space     = Channel.value("microns")
 
     // filter transcripts.parquet based on thresholds
     if ( params.filter_transcripts ) {
@@ -72,7 +73,7 @@ workflow BAYSOR_RUN_PRIOR_SEGMENTATION_MASK {
         [],
         ch_just_segmentation,
         ch_polygons2d,
-        "microns"
+        ch_coordinate_space
     )
     ch_versions = ch_versions.mix( XENIUMRANGER_IMPORT_SEGMENTATION.out.versions )
 
@@ -84,6 +85,8 @@ workflow BAYSOR_RUN_PRIOR_SEGMENTATION_MASK {
     polygons2d       = ch_polygons2d          // channel: [ ["segmentation_polygons_2d.json"] ]
     htmls            = ch_htmls               // channel: [ ["*.html"] ]
 
+    coordinate_space = ch_coordinate_space    // channel: [ "microns" ]
+
     redefined_bundle = ch_redefined_bundle    // channel: [ val(meta), "redefined-xenium-bundle" ]
 
     versions = ch_versions                    // channel: [ versions.yml ]

subworkflows/local/baysor_run_transcripts_parquet/main.nf

@@ -28,6 +28,7 @@ workflow BAYSOR_RUN_TRANSCRIPTS_PARQUET {
     ch_htmls                = Channel.empty()
 
     ch_redefined_bundle     = Channel.empty()
+    ch_coordinate_space     = Channel.value("microns")
 
 
     // generate splits
@@ -102,7 +103,7 @@ workflow BAYSOR_RUN_TRANSCRIPTS_PARQUET {
         [],
         ch_segmentation_csv,
         ch_polygons2d,
-        "microns"
+        ch_coordinate_space
     )
     ch_versions = ch_versions.mix( XENIUMRANGER_IMPORT_SEGMENTATION.out.versions )
 
@@ -114,6 +115,8 @@ workflow BAYSOR_RUN_TRANSCRIPTS_PARQUET {
     polygons2d       = ch_polygons2d          // channel: [ ["segmentation_polygons_2d.json"] ]
     htmls            = ch_htmls               // channel: [ ["*.html"] ]
 
+    coordinate_space = ch_coordinate_space    // channel: [ ["microns"] ]
+
     redefined_bundle = ch_redefined_bundle    // channel: [ val(meta), "redefined-xenium-bundle" ]
 
     versions = ch_versions                    // channel: [ versions.yml ]

subworkflows/local/cellpose_baysor_import_segmentation/main.nf

@@ -24,13 +24,15 @@ workflow CELLPOSE_BAYSOR_IMPORT_SEGMENTATION {
     ch_image                 = Channel.empty()
     ch_polygons              = Channel.empty()
     ch_segmentation          = Channel.empty()
+    ch_transcripts           = Channel.empty()
     ch_filtered_transcripts  = Channel.empty()
     ch_cellpose_cells_mask   = Channel.empty()
     ch_cellpose_nuclei_mask  = Channel.empty()
     ch_cellpose_cells_cells  = Channel.empty()
     ch_cellpose_nuclei_cells = Channel.empty()
     ch_cellpose_cells_flows  = Channel.empty()
     ch_cellpose_nuclei_flows = Channel.empty()
+    ch_coordinate_space      = Channel.value("microns")
 
     cellpose_model = params.cellpose_model ? (Channel.fromPath(params.cellpose_model, checkIfExists: true)) : []
 
@@ -98,30 +100,35 @@ workflow CELLPOSE_BAYSOR_IMPORT_SEGMENTATION {
         ch_versions = ch_versions.mix ( BAYSOR_PREPROCESS_TRANSCRIPTS.out.versions )
 
         ch_filtered_transcripts = BAYSOR_PREPROCESS_TRANSCRIPTS.out.transcripts_parquet
+        ch_transcripts = ch_filtered_transcripts
 
+    } else {
+
+        ch_transcripts = ch_transcripts_parquet
     }
 
     // run baysor with cellpose results
     if ( params.nucleus_segmentation_only ) {
 
         // run baysor with nuclei mask
-        BAYSOR_RUN ( ch_filtered_transcripts, ch_cellpose_nuclei_mask, ch_config, 30 )
+        BAYSOR_RUN ( ch_transcripts, ch_cellpose_nuclei_mask, ch_config, 30 )
         ch_versions = ch_versions.mix ( BAYSOR_RUN.out.versions )
 
     } else if ( params.cell_segmentation_only ) {
 
         // run baysor with cell mask
-        BAYSOR_RUN ( ch_filtered_transcripts, ch_cellpose_cells_mask, ch_config, 30 )
+        BAYSOR_RUN ( ch_transcripts, ch_cellpose_cells_mask, ch_config, 30 )
         ch_versions = ch_versions.mix ( BAYSOR_RUN.out.versions )
 
     } else {
 
         // run baysor with cell mask
-        BAYSOR_RUN ( ch_filtered_transcripts, [], ch_config, 30 )
+        BAYSOR_RUN ( ch_transcripts, [], ch_config, 30 )
         ch_versions = ch_versions.mix ( BAYSOR_RUN.out.versions )
 
     }
 
+
     // run import-segmentation with baysor outs
     ch_segmentation = BAYSOR_RUN.out.segmentation.map {
         _meta, segmentation -> return [ segmentation ]
@@ -136,7 +143,7 @@ workflow CELLPOSE_BAYSOR_IMPORT_SEGMENTATION {
         [],
         ch_segmentation,
         ch_polygons,
-        "microns"
+        ch_coordinate_space
     )
     ch_versions = ch_versions.mix ( XENIUMRANGER_IMPORT_SEGMENTATION.out.versions )
 
@@ -152,6 +159,8 @@ workflow CELLPOSE_BAYSOR_IMPORT_SEGMENTATION {
     segmentation_mask = ch_segmentation                            // channel: [ val(meta), [ *segmentation.csv ] ]
     polygons          = ch_polygons                                // channel: [ val(meta), [ *polygons.json ] ]
 
+    coordinate_space = ch_coordinate_space                         // channel: [ val("microns") ]
+
     redefined_bundle = XENIUMRANGER_IMPORT_SEGMENTATION.out.bundle // channel: [ val(meta), ["redefined-xenium-bundle"] ]
 
     versions         = ch_versions                                 // channel: [ versions.yml ]

subworkflows/local/cellpose_resolift_morphology_ome_tif/main.nf

@@ -24,6 +24,7 @@ workflow CELLPOSE_RESOLIFT_MORPHOLOGY_OME_TIF {
     ch_cellpose_nuclei_cells = Channel.empty()
     ch_cellpose_cells_flows  = Channel.empty()
     ch_cellpose_nuclei_flows = Channel.empty()
+    ch_coordinate_space      = Channel.value("pixels")
 
     cellpose_model = params.cellpose_model ? (Channel.fromPath(params.cellpose_model, checkIfExists: true)) : []
 
@@ -82,7 +83,7 @@ workflow CELLPOSE_RESOLIFT_MORPHOLOGY_OME_TIF {
             [],
             [],
             [],
-            ""
+            ch_coordinate_space
         )
         ch_versions = ch_versions.mix( XENIUMRANGER_IMPORT_SEGMENTATION.out.versions )
 
@@ -95,7 +96,7 @@ workflow CELLPOSE_RESOLIFT_MORPHOLOGY_OME_TIF {
             ch_cellpose_cells_mask,
             [],
             [],
-            ""
+            ch_coordinate_space
         )
         ch_versions = ch_versions.mix( XENIUMRANGER_IMPORT_SEGMENTATION.out.versions )
     }
@@ -109,6 +110,8 @@ workflow CELLPOSE_RESOLIFT_MORPHOLOGY_OME_TIF {
     nuclei_flows = ch_cellpose_nuclei_flows                        // channel: [ val(meta), [ "*flows.tif" ] ]
     nuclei_cells = ch_cellpose_nuclei_cells                        // channel: [ val(meta), [ "*seg.npy" ] ]
 
+    coordinate_space = ch_coordinate_space                         // channel: [ ["pixels"] ]
+
     redefined_bundle = XENIUMRANGER_IMPORT_SEGMENTATION.out.bundle // channel: [ val(meta), ["redefined-xenium-bundle"] ]
 
     versions = ch_versions                                         // channel: [ versions.yml ]

subworkflows/local/proseg_preset_proseg2baysor/main.nf

@@ -17,6 +17,7 @@ workflow PROSEG_PRESET_PROSEG2BAYSOR {
     main:
 
     ch_versions = Channel.empty()
+    ch_coordinate_space = Channel.value("microns")
 
     // run parquet-to-csv
     PARQUET_TO_CSV ( ch_transcripts_parquet, ".gz" )
@@ -43,7 +44,7 @@ workflow PROSEG_PRESET_PROSEG2BAYSOR {
         [],
         ch_metadata,
         ch_polygons,
-        "microns"
+        ch_coordinate_space
     )
     ch_versions = ch_versions.mix( XENIUMRANGER_IMPORT_SEGMENTATION.out.versions )
 
@@ -53,6 +54,7 @@ workflow PROSEG_PRESET_PROSEG2BAYSOR {
 
     xr_polygons           = PROSEG2BAYSOR.out.xr_polygons               // channel: [ val(meta), [ "xr-cell-polygons.geojson" ] ]
     xr_metadata           = PROSEG2BAYSOR.out.xr_metadata               // channel: [ [ "xr-transcript-metadata.csv" ] ]
+    coordinate_space      = ch_coordinate_space                         // channel: [ "microns" ]
 
     redefined_bundle      = XENIUMRANGER_IMPORT_SEGMENTATION.out.bundle // channel: [ val(meta), ["redefined-xenium-bundle"] ]
 

subworkflows/local/segger_create_train_predict/main.nf

@@ -21,6 +21,7 @@ workflow SEGGER_CREATE_TRAIN_PREDICT {
     ch_versions           = Channel.empty()
     ch_redefined_bundle   = Channel.empty()
     ch_segger_transcripts = Channel.empty()
+    ch_coordinate_space   = Channel.value("pixels")
 
     // create dataset
     SEGGER_CREATE_DATASET ( ch_basedir )
@@ -52,12 +53,13 @@ workflow SEGGER_CREATE_TRAIN_PREDICT {
         _meta, transcripts -> return [ transcripts ]
     }
 
+
     // replace transcripts.parquet in xenium bundle
-    ch_updated_bundle = ch_bundle.map { fobj ->
-        if (fobj.name == 'transcripts.parquet') {
+    ch_updated_bundle = ch_bundle.map { fileobj ->
+        if (fileobj.name == 'transcripts.parquet') {
             ch_segger_transcripts.val
         } else {
-            fobj
+            fileobj
         }
     }
 
@@ -70,7 +72,7 @@ workflow SEGGER_CREATE_TRAIN_PREDICT {
             [],
             [],
             [],
-            "pixel"
+            ch_coordinate_space
         )
         ch_redefined_bundle = XENIUMRANGER_IMPORT_SEGMENTATION.out.bundle
 
@@ -84,6 +86,8 @@ workflow SEGGER_CREATE_TRAIN_PREDICT {
     benchmarks         = SEGGER_PREDICT.out.benchmarks        // channel: [ val(meta), [ benchmarks ] ]
     segger_transcripts = ch_segger_transcripts                // channel: [ [ transcripts.parquet ] ]
 
+    coordinate_space   = ch_coordinate_space                  // channel: [ ["pixels"] ]
+
     redefined_bundle   = ch_redefined_bundle                  // channel: [ val(meta), ["redefined-xenium-bundle"] ]
 
     versions           = ch_versions                          // channel: [ versions.yml ]

subworkflows/local/spatialdata_write_meta_merge/main.nf

@@ -12,11 +12,13 @@ workflow SPATIALDATA_WRITE_META_MERGE {
     take:
     ch_bundle_path          // channel: [ val(meta), [ "path-to-xenium-bundle" ] ]
     ch_redefined_bundle     // channel: [ val(meta), [ "redefined-xenium-bundle" ] ]
+    ch_coordinate_space     // channel: [ "pixels" or "microns" ]
 
     main:
 
-    ch_versions = Channel.empty()
+    ch_versions         = Channel.empty()
     ch_segmented_object = Channel.empty()
+    ch_cells_as_circles = Channel.empty()
 
     // check segmentation - only nuclei, cells or both cells & nuclei
     if ( params.mode == 'image') {
@@ -56,7 +58,8 @@ workflow SPATIALDATA_WRITE_META_MERGE {
     SPATIALDATA_WRITE_RAW_BUNDLE (
         ch_bundle_path,
         'spatialdata_raw',
-        ch_segmented_object
+        ch_segmented_object,
+        ch_coordinate_space
     )
     ch_versions = ch_versions.mix ( SPATIALDATA_WRITE_RAW_BUNDLE.out.versions )
 
@@ -65,7 +68,8 @@ workflow SPATIALDATA_WRITE_META_MERGE {
     SPATIALDATA_WRITE_REDEFINED_BUNDLE (
         ch_redefined_bundle,
         'spatialdata_redefined',
-        ch_segmented_object
+        ch_segmented_object,
+        ch_coordinate_space
     )
     ch_versions = ch_versions.mix ( SPATIALDATA_WRITE_REDEFINED_BUNDLE.out.versions )
 

subworkflows/local/utils_nfcore_spatialxe_pipeline/main.nf

@@ -67,7 +67,7 @@ workflow PIPELINE_INITIALISATION {
     // Custom validation for pipeline parameters
     //
     validateInputParameters()
-    log.info "INFO Input params validated  ✅ "
+    log.info "INFO Pipeline parameters validated  ✅ "
 
     //
     // Create channel from input file provided through params.input
@@ -81,7 +81,7 @@ workflow PIPELINE_INITIALISATION {
         }
         .set { ch_samplesheet }
 
-        log.info "INFO Samplesheet validated   ✅ "
+        log.info "INFO Samplesheet fields validated   ✅ "
 
     } catch (Exception e) {
 
@@ -198,7 +198,7 @@ def validateInputParameters() {
     // check if segmentation mask is provided in image mode and baysor method
     if ( params.mode == 'image' && params.method == 'baysor' )
         if (!params.segmentation_mask ) {
-        log.error "❌  Error: Missing segmentation mask with `--segmentation_mask` when pipeline is run in ${params.mode} and with the ${params.method}."
+        log.warn "⚠️  Missing segmentation mask with `--segmentation_mask` when pipeline is run in ${params.mode} and with the ${params.method}. Running in coordinate mode."
     }
 
 }
@@ -268,7 +268,7 @@ def validateXeniumBundle(ch_samplesheet) {
             }
         }
     }
-    log.info "INFO Xenium bundle validated ✅ \n"
+    log.info "INFO Xenium bundle validated        ✅ \n"
 }
 
 //