feat(cellpose): update to v4.0.8 with GPU support

modules/nf-core/cellpose/environment.yml

@@ -0,0 +1,9 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - conda-forge::python=3.12
+  - pip:
+      - cellpose==4.0.8

modules/nf-core/cellpose/main.nf

@@ -1,56 +1,63 @@
 process CELLPOSE {
-    tag "$meta.id"
-    label 'process_medium'
+    tag "${meta.id}"
+    label 'process_high'
+    label 'process_gpu'
 
-    container "docker.io/biocontainers/cellpose:3.1.0_cv1"
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://community-cr-prod.seqera.io/docker/registry/v2/blobs/sha256/07/073531bf01cb425d0a869533c8751ac974c9f10b1e72c815efebdb893aa979d5/data' :
+        'community.wave.seqera.io/library/python_pip_cellpose:4be22f3ec4963ebe' }"
 
     input:
     tuple val(meta), path(image)
     path(model)
 
     output:
-    tuple val(meta), path("*masks.tif") ,   emit: mask
-    tuple val(meta), path("*flows.tif") ,   emit: flows, optional: true
-    path "versions.yml"                 ,   emit: versions
+    tuple val(meta), path("${prefix}/*masks.tif"), emit: mask
+    tuple val(meta), path("${prefix}/*flows.tif"), emit: flows, optional: true
+    tuple val(meta), path("${prefix}/*seg.npy"), emit: cells, optional: true
+    tuple val("${task.process}"), val('cellpose'), eval("pip show cellpose | grep '^Version:' | sed 's/Version: //'"), topic: versions, emit: versions_cellpose
+    tuple val("${task.process}"), val('python'), eval("python --version | sed 's/Python //'"), topic: versions, emit: versions_python
+    tuple val("${task.process}"), val('torch'), eval("pip show torch | grep '^Version:' | sed 's/Version: //'"), topic: versions, emit: versions_torch
 
     when:
     task.ext.when == null || task.ext.when
 
     script:
-    // Exit if running this module with -profile conda / -profile mamba
-    if (workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1) {
-        error "I did not manage to create a cellpose module in Conda that works in all OSes. Please use Docker / Singularity / Podman instead."
-    }
     def args = task.ext.args ?: ''
-    def model_command = model ? "--pretrained_model $model" : ""
+    def model_command = model ? "--pretrained_model ${model}" : ""
+    def gpu_flag = task.ext.use_gpu ? "--use_gpu" : ""
+    prefix = task.ext.prefix ?: "${meta.id}"
     """
     export OMP_NUM_THREADS=${task.cpus}
     export MKL_NUM_THREADS=${task.cpus}
+    # Container runs as root with HOME=/ which is not writable
+    export HOME=\$PWD
+    export MPLCONFIGDIR=\$PWD/.matplotlib
+    export CELLPOSE_LOCAL_MODELS_PATH=\$PWD/.cellpose
+    mkdir -p \$MPLCONFIGDIR \$CELLPOSE_LOCAL_MODELS_PATH
+
     cellpose \\
-        --image_path $image \\
+        --image_path ${image} \\
         --save_tif \\
-        $model_command \\
-        $args
-
-    cat <<-END_VERSIONS > versions.yml
-    "${task.process}":
-        cellpose: \$(cellpose --version | awk 'NR==2 {print \$3}')
-    END_VERSIONS
+        --verbose \\
+        ${gpu_flag} \\
+        ${model_command} \\
+        ${args}
+
+    mkdir -p ${prefix}
+    mv *masks.tif ${prefix}/
+    mv *flows.tif ${prefix}/ 2>/dev/null || true
+    mv *seg.npy ${prefix}/ 2>/dev/null || true
     """
+
     stub:
-    // Exit if running this module with -profile conda / -profile mamba
-    if (workflow.profile.tokenize(',').intersect(['conda', 'mamba']).size() >= 1) {
-        error "I did not manage to create a cellpose module in Conda that works in all OSes. Please use Docker / Singularity / Podman instead."
-    }
     def name = image.name
     def base = name.lastIndexOf('.') != -1 ? name[0..name.lastIndexOf('.') - 1] : name
-    """
-    touch ${base}_cp_masks.tif
+    prefix = task.ext.prefix ?: "${meta.id}"
 
-        cat <<-END_VERSIONS > versions.yml
-    "${task.process}":
-        cellpose: \$(cellpose --version | awk 'NR==2 {print \$3}')
-    END_VERSIONS
     """
-
+    mkdir -p ${prefix}
+    touch ${prefix}/${base}_cp_masks.tif
+    """
 }

modules/nf-core/cellpose/meta.yml

@@ -1,9 +1,11 @@
 name: "cellpose"
-description: cellpose segments cells in images
+description: cellpose segments cells in images using GPU-accelerated deep learning
 keywords:
   - segmentation
   - image
   - cellpose
+  - gpu
+  - spatial-transcriptomics
 tools:
   - "cellpose":
       description: "cellpose is an anatomical segmentation algorithm written in Python
@@ -22,13 +24,16 @@ input:
           (sample id)
     - image:
         type: file
-        description: tif file for ready for segmentation
+        description: tif file ready for segmentation
         pattern: "*.{tif,tiff}"
         ontologies: []
   - model:
       type: file
-      description: Optional input file. Cellpose 2 model trained by user using human-in-the-loop
-        approach.
+      description: |
+        Optional custom cellpose model file. When provided, passed as
+        --pretrained_model to cellpose. Pass [] (empty list) to use the
+        default model (cpsam in cellpose 4).
+      pattern: "*"
       ontologies: []
 output:
   mask:
@@ -37,32 +42,115 @@ output:
           description: |
             Groovy Map containing sample information
             [sample id]
-      - "*masks.tif":
+      - "${prefix}/*masks.tif":
           type: file
           description: labelled mask output from cellpose in tif format
-          pattern: "*.{tif, tiff}"
+          pattern: "${prefix}/*masks.tif"
           ontologies: []
   flows:
     - - meta:
           type: map
           description: |
             Groovy Map containing sample information
             [sample id]
-      - "*flows.tif":
+      - "${prefix}/*flows.tif":
           type: file
           description: cell flow output from cellpose
-          pattern: "*.{tif}"
+          pattern: "${prefix}/*flows.tif"
           ontologies: []
+  cells:
+    - - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            [sample id]
+      - "${prefix}/*seg.npy":
+          type: file
+          description: numpy array with cell segmentation data
+          pattern: "${prefix}/*seg.npy"
+          ontologies: []
+  versions_cellpose:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - cellpose:
+          type: string
+          description: The name of the tool
+      - "pip show cellpose | grep '^Version:' | sed 's/Version: //'":
+          type: eval
+          description: The expression to obtain the version of the tool
+  versions_python:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - python:
+          type: string
+          description: The name of the tool
+      - "python --version | sed 's/Python //'":
+          type: eval
+          description: The expression to obtain the version of the tool
+  versions_torch:
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - torch:
+          type: string
+          description: The name of the tool
+      - "pip show torch | grep '^Version:' | sed 's/Version: //'":
+          type: eval
+          description: The expression to obtain the version of the tool
+topics:
   versions:
-    - versions.yml:
-        type: file
-        description: File containing software versions
-        pattern: "versions.yml"
-        ontologies:
-          - edam: http://edamontology.org/format_3750 # YAML
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - cellpose:
+          type: string
+          description: The name of the tool
+      - "pip show cellpose | grep '^Version:' | sed 's/Version: //'":
+          type: eval
+          description: The expression to obtain the version of the tool
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - python:
+          type: string
+          description: The name of the tool
+      - "python --version | sed 's/Python //'":
+          type: eval
+          description: The expression to obtain the version of the tool
+    - - ${task.process}:
+          type: string
+          description: The name of the process
+      - torch:
+          type: string
+          description: The name of the tool
+      - "pip show torch | grep '^Version:' | sed 's/Version: //'":
+          type: eval
+          description: The expression to obtain the version of the tool
+notes: |
+  GPU support: Set `task.ext.use_gpu = true` to pass `--use_gpu` to cellpose.
+  The container (built via Seqera Containers) includes PyTorch 2.10.0 with CUDA 12.8
+  and falls back to CPU automatically when no GPU is available. Use the `process_gpu`
+  label to request GPU resources from your executor. When running with conda/mamba,
+  GPU support depends on having a CUDA-enabled PyTorch installation in your environment.
+
+  Model selection via the model input channel:
+    - Custom model file: file("/path/to/model")
+    - Default (cpsam):   []
+
+  Additional cellpose CLI arguments can be passed via `task.ext.args`:
+    ext.args = '--diameter 30 --flow_threshold 0.4 --cellprob_threshold 0'
+
+  Deprecated in cellpose 4.0.1+: `--chan`, `--chan2`, `--invert`, `--all_channels`,
+  `--diam_mean`, `--pretrained_model_ortho`. Do not pass these via ext.args.
+
+  Model weights are not bundled in the container. Cellpose downloads them on first
+  use to `$CELLPOSE_LOCAL_MODELS_PATH` (set to the work directory).
 authors:
   - "@josenimo"
   - "@FloWuenne"
+  - "@dongzehe"
 maintainers:
   - "@josenimo"
   - "@FloWuenne"

modules/nf-core/cellpose/tests/main.nf.test

@@ -28,7 +28,7 @@ nextflow_process {
                 { assert process.success },
                 { assert snapshot(process.out.mask).match("mask") },
                 { assert snapshot(process.out.flows).match("flows") },
-                { assert snapshot(process.out.versions).match("versions") }
+                { assert snapshot(process.out.findAll { key, val -> key.startsWith('versions') }).match("versions") }
             )
         }