Merge pull request #78 from kbestak/mindagap_skip

Mindagap skip

bin/collect_QC.py

@@ -8,6 +8,7 @@
 import pandas as pd
 from PIL import Image, ImageDraw, ImageFont
 import os
+import numpy as np
 
 
 def combine_png_files(input_paths, output_path):
@@ -85,6 +86,13 @@ def summarize_segmasks(cellxgene_table, spots_summary):
 
         ## Read in spot table
         spots = pd.read_table(args.spots, sep="\t", names=["x", "y", "z", "gene"])
+        # below code had to be added to account for the spots.txt inputs if mindagap is skipped
+        if (([val for val in spots.index.values] == [val for val in range(len(spots.index.values))]) == False):
+            spots["gene"] = spots["z"]
+            spots["z"] = spots["y"]
+            spots["y"] = spots["x"]
+            spots["x"] = spots.index
+            spots.index = range(len(spots))
         duplicated = sum(spots.gene.str.contains("Duplicated"))
         spots = spots[~spots.gene.str.contains("Duplicated")]
 

conf/modules.config

@@ -65,11 +65,12 @@ process {
         publishDir = [
             path: { "${params.outdir}/segmentation/filtered_masks" },
             mode: params.publish_dir_mode,
-            pattern: "*.{tiff,tif}"
+            pattern: "*.{tiff,tif,csv}"
         ]
     }
 
     withName: "MINDAGAP_MINDAGAP" {
+        ext.when = { !params.skip_mindagap }
         ext.args = [ "",
             params.mindagap_boxsize ? "${params.mindagap_boxsize}" : "",
             params.mindagap_loopnum ? "${params.mindagap_loopnum}" : "",
@@ -83,6 +84,7 @@ process {
     }
 
     withName: "MINDAGAP_DUPLICATEFINDER" {
+        ext.when = { !params.skip_mindagap }
         ext.args = [ "",
             params.mindagap_tilesize ? "${params.mindagap_tilesize}" : ""
         ].join(" ").trim()

nextflow.config

@@ -30,6 +30,7 @@ params {
     ilastik_multicut_project   = null
 
     // Preprocessing command line flags
+    skip_mindagap              = false
     mindagap_tilesize          = 2144
     mindagap_boxsize           = 3
     mindagap_loopnum           = 40

nextflow_schema.json

@@ -108,6 +108,11 @@
             "default": "",
             "fa_icon": "fas fa-adjust",
             "properties": {
+                "skip_mindagap": {
+                    "type": "boolean",
+                    "fa_icon": "fas fa-th-large",
+                    "description": "Skip mindagap if your data does not contain gaps between tiles."
+                },
                 "mindagap_boxsize": {
                     "type": "integer",
                     "default": 3,

subworkflows/local/utils_nfcore_molkart_pipeline/main.nf

@@ -170,7 +170,7 @@ def toolCitationText() {
     // Uncomment function in methodsDescriptionText to render in MultiQC report
     def citation_text = [
             "Tools used in the workflow included:",
-            "Mindagap (Guerreiro et al. 2023),",
+            params.skip_mindagap ? "" : "Mindagap (Guerreiro et al. 2023),",
             params.segmentation_method.split(',').contains('mesmer')   ? "Mesmer (Greenwald et al. 2021)," : "",
             params.segmentation_method.split(',').contains('ilastik')  ? "ilastik (Berg et al. 2019),"     : "",
             params.segmentation_method.split(',').contains('cellpose') ? "Cellpose (Stringer et al. 2021; Pachitariu et al 2022)," : "",
@@ -184,7 +184,7 @@ def toolBibliographyText() {
     // Can use ternary operators to dynamically construct based conditions, e.g. params["run_xyz"] ? "<li>Author (2023) Pub name, Journal, DOI</li>" : "",
     // Uncomment function in methodsDescriptionText to render in MultiQC report
     def reference_text = [
-        "<li>Ricardo Guerreiro, Florian Wuennemann, & pvtodorov. (2023). ViriatoII/MindaGap: v0.0.3 (0.0.3). Zenodo. https://doi.org/10.5281/zenodo.8120559",
+        params.skip_mindagap ? "" : "<li>Ricardo Guerreiro, Florian Wuennemann, & pvtodorov. (2023). ViriatoII/MindaGap: v0.0.3 (0.0.3). Zenodo. https://doi.org/10.5281/zenodo.8120559",
         params.segmentation_method.split(',').contains('mesmer')   ? "<li>Greenwald, N.F., Miller, G., Moen, E. et al. Whole-cell segmentation of tissue images with human-level performance using large-scale data annotation and deep learning. Nat Biotechnol 40, 555–565 (2022). https://doi.org/10.1038/s41587-021-01094-0</li>" : "",
         params.segmentation_method.split(',').contains('ilastik')  ? "<li>Berg, S., Kutra, D., Kroeger, T. et al. ilastik: interactive machine learning for (bio)image analysis. Nat Methods 16, 1226–1232 (2019). https://doi.org/10.1038/s41592-019-0582-9</li>" : "",
         params.segmentation_method.split(',').contains('cellpose') ? "<li>Stringer, C., Wang, T., Michaelos, M. et al. Cellpose: a generalist algorithm for cellular segmentation. Nat Methods 18, 100–106 (2021). https://doi.org/10.1038/s41592-020-01018-x</li>" : "",

tests/main.nf.test

@@ -78,6 +78,32 @@ nextflow_pipeline {
         }
     }
 
+    test("Skip mindagap - clahe - cellpose") {
+
+        when {
+            params {
+                input = 'https://raw.githubusercontent.com/nf-core/test-datasets/molkart/test_data/samplesheets/samplesheet_membrane.csv'
+                outdir = "$outputDir"
+                clahe_pyramid_tile = 368
+                segmentation_method = "cellpose"
+                skip_mindagap = true
+                }
+        }
+
+        then {
+            assert workflow.success
+            assert snapshot(
+                path("$outputDir/segmentation/cellpose/mem_only_cellpose_mask.tif"),
+                path("$outputDir/segmentation/filtered_masks/mem_only_cellpose_filtered.tif"),
+                path("$outputDir/spot2cell/cellxgene_mem_only_cellpose.csv"),
+                path("$outputDir/anndata/mem_only_cellpose.adata"),
+                path("$outputDir/molkartqc/mem_only.cellpose.spot_QC.csv"),
+            ).match()
+            assert file("$outputDir/stack/mem_only_stack.ome.tif").exists()
+            assert file("$outputDir/multiqc/multiqc_report.html").exists()
+        }
+    }
+
     test("Create training subset") {
 
         when {

tests/main.nf.test.snap

@@ -42,6 +42,20 @@
         },
         "timestamp": "2024-06-11T11:47:51.185058276"
     },
+    "Skip mindagap - clahe - cellpose": {
+        "content": [
+            "mem_only_cellpose_mask.tif:md5,8faad3e707aba96d3b93d9419311e843",
+            "mem_only_cellpose_filtered.tif:md5,5c566683d7af62721415329799ad2ac1",
+            "cellxgene_mem_only_cellpose.csv:md5,925f5fabeeafc520fe839cabc79fa9ce",
+            "mem_only_cellpose.adata:md5,a0aa767a32929417f7f3253d12278f3a",
+            "mem_only.cellpose.spot_QC.csv:md5,e393966b17effbc05c1b93bb590d7e80"
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "24.04.4"
+        },
+        "timestamp": "2024-09-10T15:07:17.243436747"
+    },
     "Create training subset": {
         "content": [
             "nuc_only_nuclear_gridfilled.tiff:md5,123763d54f05b2274690b3b84f9690b1"

tests/nextflow.config

@@ -4,3 +4,4 @@
 ========================================================================================
 */
 
+trace.overwrite = true

workflows/molkart.nf

@@ -82,10 +82,11 @@ workflow MOLKART {
     // MODULE: Apply Contrast-limited adaptive histogram equalization (CLAHE)
     // CLAHE is either applied to all images, or none.
     //
-    CLAHE(MINDAGAP_MINDAGAP.out.tiff)
+    clahe_in = params.skip_mindagap ? mindagap_in : MINDAGAP_MINDAGAP.out.tiff
+    CLAHE(clahe_in)
     ch_versions = ch_versions.mix(CLAHE.out.versions)
 
-    map_for_stacks = !params.skip_clahe ? CLAHE.out.img_clahe : MINDAGAP_MINDAGAP.out.tiff
+    map_for_stacks = params.skip_clahe ? clahe_in : CLAHE.out.img_clahe
 
     map_for_stacks
         .map {
@@ -148,8 +149,7 @@ workflow MOLKART {
     MINDAGAP_DUPLICATEFINDER(spot_tuple)
     ch_versions = ch_versions.mix(MINDAGAP_DUPLICATEFINDER.out.versions)
 
-    qc_spots = MINDAGAP_DUPLICATEFINDER.out.marked_dups_spots
-
+    qc_spots = params.skip_mindagap ? spot_tuple : MINDAGAP_DUPLICATEFINDER.out.marked_dups_spots
     //
     // MODULE: DeepCell Mesmer segmentation
     //
@@ -278,7 +278,6 @@ workflow MOLKART {
     qc_spots
         .combine(spot2cell_out, by: 0)
         .set{ molkartqc }
-
     MOLKARTQC(molkartqc)
     ch_versions = ch_versions.mix(MOLKARTQC.out.versions)