Build branch openpipeline_spatial/niche-compass with version niche-compass to openpipeline_spatial on branch niche-compass (87e6260)

Build pipeline: openpipelines-bio.openpipeline-spatial.niche-compass-8zcq9 Source commit: 87e62605aa Source message: add cmake to test container
2026-03-25 11:11:31 +00:00
parent f057217658
commit adbd282fec
393 changed files with 26718 additions and 1455 deletions
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,10 +1,24 @@
-# openpipeline_spatial x.x.x
+# openpipeline_spatial 0.4.0
+
+## NEW FUNCTIONALITY
+
+* `feature_annotation/spatial_autocorr`: Added a spatial autocorrelation component (PR #45).
+
+* `feature_annotation/xenium_spatial_statistics`: Added a component to calculate spatial statistics on Xenium-ingested data (PR #46).
+
+* `neighbors/join_graphs`: Combine spatial and expression neighborhood graphs into a single graph (PR #47).
+
+# openpipeline_spatial 0.3.0
+
+## NEW FUNCTIONALITY
+
+* `convert/from_h5mu_to_spatialdata`: Added a converter component to convert from H5MU to SpatialData (PR #40)

 ## MINOR CHANGES

 * Bump squidpy to 1.8.1 and spatialdata to 0.7.2 (PR #41).

-* Update openpipeline dependencies to v4.0.2 (PR #42).
+* Update openpipeline dependencies to v4.0.3 (PR #42, PR #44).

 # openpipeline_spatial 0.2.0

--- a/_viash.yaml
+++ b/_viash.yaml
@@ -10,7 +10,7 @@ repositories:
  - name: openpipeline
    repo: openpipeline
    type: vsh
-    tag: v4.0.2
+    tag: v4.0.3
 info:
  test_resources:
    - type: s3
--- a/resources_test_scripts/visium_tiny.sh
+++ b/resources_test_scripts/visium_tiny.sh
@@ -7,6 +7,7 @@ REPO_ROOT=$(git rev-parse --show-toplevel)

 # Define absolute directory path
 DIR="$REPO_ROOT/resources_test/visium"
+ID="Visium_FFPE_Human_Ovarian_Cancer_tiny"

 # from https://www.10xgenomics.com/resources/datasets/human-ovarian-cancer-1-standard
 mkdir -p "$DIR"
@@ -46,6 +47,10 @@ echo "> Running spaceranger complete"
 rm -rf "$DIR/Visium_FFPE_Human_Ovarian_Cancer_fastqs"
 rm -f "$DIR/Visium_FFPE_Human_Ovarian_Cancer_image.jpg"

+viash run "$REPO_ROOT/src/convert/from_spaceranger_to_h5mu/config.vsh.yaml" -- \
+    --input "$DIR/Visium_FFPE_Human_Ovarian_Cancer_tiny_spaceranger" \
+    --output "$DIR/$ID.h5mu"
+
 aws s3 sync \
    --profile di \
    --exclude "*.yaml" \
--- a/resources_test_scripts/xenium_tiny.sh
+++ b/resources_test_scripts/xenium_tiny.sh
@@ -19,14 +19,16 @@ function clean_up {
 trap clean_up EXIT

 if [ ! -d "$OUT" ]; then
-    tiny_dataset="https://raw.githubusercontent.com/nf-core/test-datasets/spatialxe/Xenium_Prime_Mouse_Ileum_tiny_outs.zip"
-    wget "$tiny_dataset" -O "$TMPDIR/xenium_tiny.zip"
+    tiny_dataset="https://raw.githubusercontent.com/nf-core/test-datasets/spatialxe/Xenium_Prime_Mouse_Ileum_tiny_outs.tar.gz"
+    wget "$tiny_dataset" -O "$TMPDIR/xenium_tiny.tar.gz"

-    unzip -q "$TMPDIR/xenium_tiny.zip" -d "$TMPDIR/xenium_tiny"
+    mkdir -p "$TMPDIR/xenium_tiny"
+    tar -xzf "$TMPDIR/xenium_tiny.tar.gz" -C "$TMPDIR/xenium_tiny"
    mkdir -p "$OUT"
    mv "$TMPDIR/xenium_tiny/Xenium_Prime_Mouse_Ileum_tiny_outs/"* "$OUT/"
 fi

+rm -rf "$DIR/$ID.zarr"
 viash run "$REPO_ROOT/src/convert/from_xenium_to_spatialdata/config.vsh.yaml" -- \
    --input "$OUT" \
    --output "$DIR/$ID.zarr"
@@ -35,11 +37,73 @@ viash run "$REPO_ROOT/src/convert/from_spatialdata_to_h5mu/config.vsh.yaml" -- \
    --input "$DIR/$ID.zarr" \
    --output "$DIR/$ID.h5mu"

-
 viash run src/neighbors/spatial_neighborhood_graph/config.vsh.yaml -- \
    --input "$DIR/$ID.h5mu" \
    --output "$DIR/$ID.h5mu"

+cat > /tmp/qc.yaml <<EOF
+param_list:
+  - id: xenium_tiny
+    input: "$DIR/$ID.h5mu"
+var_name_mitochondrial_genes: mitochondrial
+var_name_ribosomal_genes: ribosomal
+output: '\$id.qc.h5mu'
+output_compression: gzip
+publish_dir: "$DIR"
+EOF
+
+nextflow run openpipelines-bio/openpipeline \
+  -latest \
+  -r 2.1.0 \
+  -main-script target/nextflow/workflows/qc/qc/main.nf \
+  -profile docker \
+  -params-file /tmp/qc.yaml \
+  -resume \
+  -config src/workflows/utils/labels_ci.config
+
+viash run "$REPO_ROOT/src/neighbors/spatial_neighborhood_graph/config.vsh.yaml" -- \
+    --input "$DIR/$ID.qc.h5mu" \
+    --output "$DIR/${ID}.qc.neighbors.h5mu"
+
+cat > /tmp/pca.yaml <<EOF
+param_list:
+  - id: xenium_tiny
+    input: "$DIR/${ID}.qc.neighbors.h5mu"
+output: '\$id.qc.neighbors.pca.h5mu'
+output_compression: gzip
+publish_dir: "$TMPDIR"
+EOF
+
+nextflow run openpipelines-bio/openpipeline \
+  -latest \
+  -r v4.0.3 \
+  -main-script target/nextflow/dimred/pca/main.nf \
+  -profile docker \
+  -params-file /tmp/pca.yaml \
+  -config src/workflows/utils/labels_ci.config \
+  -resume
+
+cat > /tmp/find_neighbors.yaml <<EOF
+param_list:
+  - id: xenium_tiny
+    input: "$TMPDIR/xenium_tiny.qc.neighbors.pca.h5mu"
+output: '\$id.qc.all_neighbors.pca.h5mu'
+output_compression: gzip
+publish_dir: "$TMPDIR"
+EOF
+
+nextflow run openpipelines-bio/openpipeline \
+  -latest \
+  -r v4.0.3 \
+  -main-script target/nextflow/neighbors/find_neighbors/main.nf \
+  -profile docker \
+  -params-file /tmp/find_neighbors.yaml \
+  -config src/workflows/utils/labels_ci.config \
+  -resume
+
+# Move the final output to the destination directory
+mv "$TMPDIR/xenium_tiny.qc.all_neighbors.pca.h5mu" "$DIR/xenium_tiny.qc.all_neighbors.pca.h5mu"
+
 # Sync to S3
 aws s3 sync \
    --profile di \
--- a/src/authors/luke_zappia.yaml
+++ b/src/authors/luke_zappia.yaml
@@ -0,0 +1,12 @@
+name: Luke Zappia
+info:
+  role: Contributor
+  links:
+    email: luke@data-intuitive.com
+    github: lazappi
+    orcid: 0000-0001-7744-8565
+    linkedin: lazappi
+  organizations:
+    - name: Data Intuitive
+      href: https://www.data-intuitive.com
+      role: Data Science Engineer
--- a/src/base/requirements/spatialdata.yaml
+++ b/src/base/requirements/spatialdata.yaml
@@ -1,3 +1,6 @@
 packages:
  - spatialdata~=0.7.2
+  - ome-zarr~=0.12.2
  - pyarrow~=18.0.0
+script: |
+    exec("try:\n  import zarr; from importlib.metadata import version\nexcept ModuleNotFoundError:\n  exit(0)\nelse:  assert int(version(\"zarr\").partition(\".\")[0]) > 2")
--- a/src/convert/from_h5mu_to_spatialdata/config.vsh.yaml
+++ b/src/convert/from_h5mu_to_spatialdata/config.vsh.yaml
@@ -0,0 +1,64 @@
+name: "from_h5mu_to_spatialdata"
+namespace: "convert"
+scope: "public"
+description: |
+  Reads in an H5MU file and saves it as a SpatialData Zarr store. The selected
+  modality in the MuData is stored as the main table in the SpatialData object.
+  If a matching existing SpatialData is provided, it will be used to fill the
+  remaining SpatialData slots.
+    
+authors:
+  - __merge__: /src/authors/dorien_roosen.yaml
+    roles: [ maintainer ]
+  - __merge__: /src/authors/luke_zappia.yaml
+    roles: [ author ]
+arguments:
+  - name: "--input"
+    alternatives: ["-i"]
+    type: file
+    description: Input H5MU file.
+    example: input.h5mu
+    direction: input
+    required: true
+  - name: "--input_spatialdata"
+    type: file
+    description: An optional existing SpatialData Zarr store to fill remaining slots from.
+    example: existing.zarr
+    direction: input
+    required: false
+  - name: "--modality"
+    type: string
+    default: rna
+    description: The modality in the MuData to be used as the main table in the SpatialData object.
+  - name: "--output"
+    alternatives: ["-o"]
+    type: file
+    description: The path to the output SpatialData Zarr store.
+    example: "output.zarr"
+    direction: output
+    required: true
+
+resources:
+  - type: python_script
+    path: script.py
+  - path: /src/utils/setup_logger.py
+test_resources:
+  - type: python_script
+    path: test.py
+  - path: /resources_test/xenium/xenium_tiny.h5mu
+  - path: /resources_test/xenium/xenium_tiny.zarr
+engines:
+  - type: docker
+    image: python:3.12-slim
+    setup:
+      - type: apt
+        packages: 
+          - procps
+      - type: python
+        __merge__: [/src/base/requirements/anndata_mudata.yaml, /src/base/requirements/spatialdata.yaml]
+    __merge__: [ /src/base/requirements/python_test_setup.yaml, .]
+runners:
+  - type: executable
+  - type: nextflow
+    directives:
+      label: [lowmem, singlecpu]
--- a/src/convert/from_h5mu_to_spatialdata/script.py
+++ b/src/convert/from_h5mu_to_spatialdata/script.py
@@ -0,0 +1,60 @@
+import logging
+import sys
+
+import mudata as mu
+import spatialdata as sd
+
+## VIASH START
+par = {
+    "input": "./resources_test/xenium/xenium_tiny.h5mu",
+    "input_spatialdata": "./resources_test/xenium/xenium_tiny.zarr",
+    "output": "./resources_test/xenium/xenium_tiny_from_h5mu.zarr",
+    "modality": "rna",
+}
+meta = {"resources_dir": "src/utils"}
+## VIASH END
+
+sys.path.append(meta["resources_dir"])
+from setup_logger import setup_logger
+
+logger = setup_logger()
+
+logger.info("Starting conversion from H5MU to SpatialData...")
+
+logger.info(f"Reading input H5MU file from {par['input']}...")
+mdata = mu.read_h5mu(par["input"])
+
+logger.info("Extracting modality from MuData object...")
+mod = mdata.mod[par["modality"]]
+
+if par.get("input_spatialdata", None) is not None:
+    logger.info(f"Reading existing SpatialData from {par['input_spatialdata']}...")
+
+    # Disable logger messages from spatialdata when reading
+    logger.setLevel(logging.WARNING)
+    sdata_existing = sd.read_zarr(par["input_spatialdata"])
+    logger.setLevel(logging.INFO)
+
+    logger.info("Checking modality matches existing SpatialData table...")
+    if not mod.n_obs == sdata_existing["table"].n_obs:
+        raise ValueError(
+            "The number of observations in the selected modality does not match the existing SpatialData table."
+        )
+    if not mod.obs_names.equals(sdata_existing["table"].obs_names):
+        raise ValueError(
+            "The observation names in the selected modality do not match the existing SpatialData table."
+        )
+
+logger.info("Creating SpatialData object...")
+if par.get("input_spatialdata", None) is not None:
+    logger.info("Using existing SpatialData...")
+    sdata = sdata_existing
+    sdata["table"] = mod
+else:
+    logger.info("Creating new SpatialData...")
+    sdata = sd.SpatialData(tables={"table": mod})
+
+logger.info(f"Writing output SpatialData Zarr store to {par['output']}...")
+sdata.write(par["output"], overwrite=True)
+
+logger.info("Done!")
--- a/src/convert/from_h5mu_to_spatialdata/test.py
+++ b/src/convert/from_h5mu_to_spatialdata/test.py
@@ -0,0 +1,143 @@
+import os
+import sys
+
+import mudata as mu
+import pytest
+import spatialdata as sd
+
+
+def test_simple_execution(run_component, tmp_path):
+    input = meta["resources_dir"] + "/xenium_tiny.h5mu"
+    input_spatialdata = meta["resources_dir"] + "/xenium_tiny.zarr"
+    output = tmp_path / "output.zarr"
+
+    run_component(
+        [
+            "--input",
+            input,
+            "--input_spatialdata",
+            input_spatialdata,
+            "--output",
+            output,
+        ]
+    )
+    assert os.path.exists(output), "output zarr was not created"
+
+    mdata = mu.read_h5mu(input)
+    mod = mdata.mod["rna"]
+
+    sdata = sd.read_zarr(output)
+    table = sdata["table"]
+
+    # Check that the main table in the SpatialData object matches the selected modality in the MuData object
+    assert table.n_obs == mod.n_obs, (
+        "The number of observations in the SpatialData table does not match the selected modality in the MuData object."
+    )
+    assert table.obs_names.equals(mod.obs_names), (
+        "The observation names in the SpatialData table do not match the selected modality in the MuData object."
+    )
+    assert table.n_vars == mod.n_vars, (
+        "The number of variables in the SpatialData table does not match the selected modality in the MuData object."
+    )
+    assert table.var_names.equals(mod.var_names), (
+        "The variable names in the SpatialData table do not match the selected modality in the MuData object."
+    )
+    assert table.obs.keys().equals(mod.obs.keys()), (
+        "The observation metadata columns in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert table.var.keys().equals(mod.var.keys()), (
+        "The variable metadata columns in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert set(table.uns.keys()) == set(mod.uns.keys()), (
+        "The unstructured metadata keys in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert set(table.obsm.keys()) == set(mod.obsm.keys()), (
+        "The obsm keys in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert set(table.varm.keys()) == set(mod.varm.keys()), (
+        "The varm keys in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+
+    sdata_existing = sd.read_zarr(input_spatialdata)
+
+    # Check that remaining slots in the SpatialData object are filled from the existing SpatialData when provided
+    assert sdata.images.keys() == sdata_existing.images.keys(), (
+        "The image keys in the output SpatialData do not match those in the existing SpatialData."
+    )
+    assert sdata.labels.keys() == sdata_existing.labels.keys(), (
+        "The label keys in the output SpatialData do not match those in the existing SpatialData."
+    )
+    assert sdata.shapes.keys() == sdata_existing.shapes.keys(), (
+        "The shape keys in the output SpatialData do not match those in the existing SpatialData."
+    )
+    assert sdata.points.keys() == sdata_existing.points.keys(), (
+        "The point keys in the output SpatialData do not match those in the existing SpatialData."
+    )
+
+
+def test_execution_without_input_spatialdata(run_component, tmp_path):
+    input = meta["resources_dir"] + "/xenium_tiny.h5mu"
+    output = tmp_path / "output_no_spatialdata.zarr"
+
+    run_component(
+        [
+            "--input",
+            input,
+            "--output",
+            output,
+        ]
+    )
+    assert os.path.exists(output), "output zarr was not created"
+
+    mdata = mu.read_h5mu(input)
+    mod = mdata.mod["rna"]
+
+    sdata = sd.read_zarr(output)
+    table = sdata["table"]
+
+    # Check that the main table in the SpatialData object matches the selected modality in the MuData object
+    assert table.n_obs == mod.n_obs, (
+        "The number of observations in the SpatialData table does not match the selected modality in the MuData object."
+    )
+    assert table.obs_names.equals(mod.obs_names), (
+        "The observation names in the SpatialData table do not match the selected modality in the MuData object."
+    )
+    assert table.n_vars == mod.n_vars, (
+        "The number of variables in the SpatialData table does not match the selected modality in the MuData object."
+    )
+    assert table.var_names.equals(mod.var_names), (
+        "The variable names in the SpatialData table do not match the selected modality in the MuData object."
+    )
+    assert table.obs.keys().equals(mod.obs.keys()), (
+        "The observation metadata columns in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert table.var.keys().equals(mod.var.keys()), (
+        "The variable metadata columns in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert set(table.uns.keys()) == set(mod.uns.keys()), (
+        "The unstructured metadata keys in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert set(table.obsm.keys()) == set(mod.obsm.keys()), (
+        "The obsm keys in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+    assert set(table.varm.keys()) == set(mod.varm.keys()), (
+        "The varm keys in the SpatialData table do not match those in the selected modality of the MuData object."
+    )
+
+    # When no existing SpatialData is provided, remaining slots should be empty
+    assert len(sdata.images) == 0, (
+        "Expected no images when input_spatialdata is not provided."
+    )
+    assert len(sdata.labels) == 0, (
+        "Expected no labels when input_spatialdata is not provided."
+    )
+    assert len(sdata.shapes) == 0, (
+        "Expected no shapes when input_spatialdata is not provided."
+    )
+    assert len(sdata.points) == 0, (
+        "Expected no points when input_spatialdata is not provided."
+    )
+
+
+if __name__ == "__main__":
+    sys.exit(pytest.main([__file__]))
--- a/src/convert/from_h5mu_to_spatialexperiment/config.vsh.yaml
+++ b/src/convert/from_h5mu_to_spatialexperiment/config.vsh.yaml
@@ -64,6 +64,7 @@ engines:
          - python3-pip
          - python3-dev
          - python-is-python3
+          - cmake
      - type: r
        cran: [ reticulate, testthat ]
      - type: python
--- a/src/convert/from_xenium_to_spatialexperiment/config.vsh.yaml
+++ b/src/convert/from_xenium_to_spatialexperiment/config.vsh.yaml
@@ -68,8 +68,6 @@ engines:
        env: ["LIBARROW_MINIMAL=false"]
      - type: r
        script: 'install.packages("arrow", type = "source")'
-      - type: r
-        script: 'Sys.setenv(LIBARROW_MINIMAL = "false"); install.packages("arrow", type = "source")'
      - type: r
        bioc: [ SpatialExperimentIO ]
    test_setup:
--- a/src/dataflow/split_h5mu/config.vsh.yaml
+++ b/src/dataflow/split_h5mu/config.vsh.yaml
@@ -1,73 +0,0 @@
-name: split_h5mu
-namespace: "dataflow"
-scope: "public"
-description: |
-  Split the samples of a single modality from a .h5mu (multimodal) sample into seperate .h5mu files based on the values of an .obs column of this modality. 
-authors:
-  - __merge__: /src/authors/dorien_roosen.yaml
-    roles: [ author, maintainer ]
-argument_groups:
- name: Input & specifications
-  arguments:
-  - name: "--input"
-    type: file
-    description: Path to a single .h5mu file.
-    required: true
-  - name: "--modality"
-    description: |
-      Which modality from the input MuData file to process.
-    type: string
-    default: "rna"
-    required: false
-  - name: "--obs_feature"
-    type: string
-    required: true
-    description: The .obs column to split the mudata on.
-    example: "celltype"
-  - name: "--drop_obs_nan"
-    type: boolean_true
-    description: Whether to drop all .obs columns that contain only nan values after splitting.
-  - name: "--ensure_unique_filenames"
-    type: boolean_true
-    description: Append number suffixes to ensure unique filenames after sanitizing obs feature values.
-
- name: Outputs
-  arguments:
-  - name: "--output"
-    type: file
-    required: true
-    direction: output
-    example: "/path/to/output"
-    description: Output directory containing multiple h5mu files.
-  - name: "--output_files"
-    type: file
-    required: true
-    direction: output
-    example: sample_files.csv
-    description: A csv containing the base filename and obs feature by which it was split.
-  __merge__: [., /src/base/h5_compression_argument.yaml]
-
-
-resources:
-  - type: python_script
-    path: script.py
-  - path: /src/utils/setup_logger.py
-test_resources:
-  - type: python_script
-    path: test.py
-
-engines:
-  - type: docker
-    image: python:3.12-slim
-    setup:
-      - type: apt
-        packages: 
-          - procps
-      - type: python
-        __merge__: /src/base/requirements/anndata_mudata.yaml
-    __merge__: [ /src/base/requirements/python_test_setup.yaml, .]
-runners:
-  - type: executable
-  - type: nextflow
-    directives:
-      label: [ lowcpu, highmem, highdisk]
--- a/src/dataflow/split_h5mu/script.py
+++ b/src/dataflow/split_h5mu/script.py
@@ -1,119 +0,0 @@
-import sys
-import mudata as mu
-import pandas as pd
-import re
-import gc
-from pathlib import Path
-from collections import defaultdict
-
-### VIASH START
-par = {
-    "input": "cart_atomx_process_samples.h5mu",
-    "modality": "rna",
-    "obs_feature": "sample_id",
-    "output": "reference_download/sample_split",
-    "drop_obs_nan": "true",
-    "output_compression": None,
-    "output_files": "reference_download/sample_files.csv",
-    "ensure_unique_filenames": True,
-}
-meta = {
-    "resources_dir": "src/utils",
-}
-# import anndata as ad
-
-# df = pd.DataFrame(
-#     [[1, 2, 3], [4, 5, 6]], index=["obs1", "obs2"], columns=["var1", "var2", "var3"]
-# )
-# var3 = pd.DataFrame(["d", "e", "g"], index=df.columns, columns=["Feat"])
-# obs3 = pd.DataFrame(["C C", "C_C"], index=df.index, columns=["Obs"])
-# ad3 = ad.AnnData(df, obs=obs3, var=var3)
-# mdata = mu.MuData({"rna": ad3})
-# mdata.write_h5mu("test_san.h5mu")
-# par["input"] = "test_san.h5mu"
-# par["obs_feature"] = "Obs"
-### VIASH END
-
-sys.path.append(meta["resources_dir"])
-from setup_logger import setup_logger
-
-logger = setup_logger()
-
-
-def main():
-    logger.info(f"Reading {par['input']}")
-    input_file = Path(par["input"].strip())
-
-    mdata = mu.read_h5mu(input_file)
-    adata = mdata.mod[par["modality"]]
-
-    logger.info(f"Reading unique features from {par['obs_feature']}")
-    obs_features = adata.obs[par["obs_feature"]].unique().tolist()
-
-    # sanitize --obs_feature values
-    obs_features_s = [re.sub(r"[-\s]", "_", str(s).strip()) for s in obs_features]
-    obs_features_s = [re.sub(r"[^A-Za-z0-9_]", "", s) for s in obs_features_s]
-
-    # ensure that names are unique, if not raise or append number as suffix
-    if not len(obs_features_s) == len(set(obs_features_s)):
-        if not par["ensure_unique_filenames"]:
-            raise ValueError(
-                f"File names are not unique after sanitizing the --obs_feature {par['obs_feature']} values"
-            )
-
-        logger.info("Ensuring unique names for par['obs_feature']")
-        counts = defaultdict(lambda: -1)
-        for i, feature in enumerate(obs_features_s):
-            counts[feature] += 1
-            if (curr_counts := counts[feature]) > 0:
-                obs_features_s[i] += f"_{curr_counts}"
-
-    # generate output dir
-    output_dir = Path(par["output"])
-    if not output_dir.is_dir():
-        output_dir.mkdir(parents=True)
-
-    # split modality of mdata file base on obs_feature
-    logger.info(f"Splitting file based on {par['obs_feature']} values {obs_features}")
-    obs_files = []
-
-    for obs_name, file_name in zip(obs_features, obs_features_s):
-        logger.info(
-            f"Filtering modality '{par['modality']}' observations by .obs['{par['obs_feature']}'] == {obs_name}"
-        )
-        mdata_obs = mdata.copy()
-        adata_full = mdata_obs.mod[par["modality"]]
-
-        # split the samples
-        mask = adata_full.obs[par["obs_feature"]] == obs_name
-        adata_obs = adata_full[mask].copy()
-
-        # Dropping columns that only have nan values after splitting
-        if par["drop_obs_nan"]:
-            logger.info("Dropping all .obs columns with NaN values")
-            adata_obs.obs = adata_obs.obs.dropna(axis=1, how="all")
-
-        mdata_obs.mod[par["modality"]] = adata_obs
-
-        mdata_obs_name = f"{input_file.stem}_{file_name}.h5mu"
-        out_path = output_dir / mdata_obs_name
-
-        # replace mdata file with modality adata contianing split samples
-        logger.info(
-            f"Writing h5mu filtered for {par['obs_feature']} {obs_name} to file {out_path}"
-        )
-
-        mdata_obs.write_h5mu(out_path, compression=par["output_compression"])
-
-        # avoid keeping files in memory
-        obs_files.append(mdata_obs_name)
-        del mdata_obs, adata_obs
-        gc.collect()
-
-    logger.info(f"Writing output_files CSV file to {par['output_files']}")
-    df = pd.DataFrame({"name": obs_features_s, "filename": obs_files})
-    df.to_csv(par["output_files"], index=False)
-
-
-if __name__ == "__main__":
-    main()
--- a/src/dataflow/split_h5mu/test.py
+++ b/src/dataflow/split_h5mu/test.py
@@ -1,288 +0,0 @@
-import sys
-from textwrap import dedent
-import pytest
-import mudata as mu
-import anndata as ad
-import numpy as np
-import pandas as pd
-import subprocess
-import re
-
-
-@pytest.fixture
-def input_modality_1():
-    df = pd.DataFrame(
-        [[1, 2, 3], [4, 5, 6]], index=["obs1", "obs2"], columns=["var1", "var2", "var3"]
-    )
-    obs = pd.DataFrame({"Obs": ["A", "B"], "Obs_nan": [np.nan, np.nan]}, index=df.index)
-    var = pd.DataFrame([["a"], ["b"], ["c"]], index=df.columns, columns=["Feat"])
-    ad1 = ad.AnnData(df, obs=obs, var=var)
-    return ad1
-
-
-@pytest.fixture
-def input_modality_2():
-    df = pd.DataFrame(
-        [[1, 2, 3], [4, 5, 6]], index=["obs1", "obs2"], columns=["var1", "var2", "var3"]
-    )
-    var2 = pd.DataFrame(["d", "e", "g"], index=df.columns, columns=["Feat"])
-    obs2 = pd.DataFrame(["C", "D"], index=df.index, columns=["Obs"])
-    ad2 = ad.AnnData(df, obs=obs2, var=var2)
-    return ad2
-
-
-@pytest.fixture
-def input_modality_3():
-    df = pd.DataFrame(
-        [[1, 2, 3], [4, 5, 6]], index=["obs1", "obs2"], columns=["var1", "var2", "var3"]
-    )
-    var3 = pd.DataFrame(["d", "e", "g"], index=df.columns, columns=["Feat"])
-    obs3 = pd.DataFrame(["C C", "C_C"], index=df.index, columns=["Obs"])
-    ad3 = ad.AnnData(df, obs=obs3, var=var3)
-    return ad3
-
-
-@pytest.fixture
-def input_h5mu(input_modality_1, input_modality_2):
-    tmp_mudata = mu.MuData({"mod1": input_modality_1, "mod2": input_modality_2})
-    return tmp_mudata
-
-
-@pytest.fixture
-def input_h5mu_path(write_mudata_to_file, input_h5mu):
-    return write_mudata_to_file(input_h5mu)
-
-
-@pytest.fixture
-def input_h5mu_non_unique_filenames(input_modality_3):
-    tmp_mudata = mu.MuData({"mod3": input_modality_3})
-    return tmp_mudata
-
-
-@pytest.fixture
-def input_h5mu_path_non_unique_filenames(
-    write_mudata_to_file, input_h5mu_non_unique_filenames
-):
-    return write_mudata_to_file(input_h5mu_non_unique_filenames)
-
-
-def test_sample_split(run_component, random_path, input_h5mu, input_h5mu_path):
-    output_dir = random_path()
-    output_files = random_path(extension="csv")
-    args = [
-        "--input",
-        input_h5mu_path,
-        "--output",
-        str(output_dir),
-        "--modality",
-        "mod1",
-        "--obs_feature",
-        "Obs",
-        "--output_files",
-        str(output_files),
-    ]
-
-    run_component(args)
-    assert output_files.is_file()
-    assert output_dir.is_dir()
-
-    # check output dir and file names
-    dir_content = [
-        h5mu_file
-        for h5mu_file in output_dir.iterdir()
-        if h5mu_file.suffix == ".h5mu" and h5mu_file != input_h5mu_path
-    ]
-    s1_file = output_dir / f"{input_h5mu_path.stem}_A.h5mu"
-    s2_file = output_dir / f"{input_h5mu_path.stem}_B.h5mu"
-    assert set(dir_content) == set([s1_file, s2_file])
-
-    # check that number of modalities, variables and observations
-    s1 = mu.read_h5mu(s1_file)
-    s2 = mu.read_h5mu(s2_file)
-    assert s1.n_mod == 2
-    assert s2.n_mod == 2
-
-    assert s1.n_obs == input_h5mu.n_obs, (
-        "number of observations of split file does not match input file"
-    )
-    assert s2.n_obs == input_h5mu.n_obs, (
-        "number of observations of split file does not match input file"
-    )
-
-    assert s1.mod["mod1"].n_obs == 1, (
-        "number of observations of split file s1 modality mod1 should equal 1"
-    )
-    assert s1.mod["mod2"].n_obs == input_h5mu.n_obs, (
-        "number of observations of split file s1 modality mod2 should equal input file"
-    )
-
-    assert len(s1.mod["mod1"].obs.keys()) == 2, (
-        "number of observation keys split file s1 modality mod1 should equal 2"
-    )
-    assert len(s1.mod["mod2"].obs.keys()) == 1, (
-        "number of observation keys split file s1 modality mod2 should equal 1"
-    )
-
-    assert s2.mod["mod1"].n_obs == 1, (
-        "number of observations of split file s2 modality mod1 should equal 1"
-    )
-    assert s2.mod["mod2"].n_obs == input_h5mu.n_obs, (
-        "number of observations of split file s2 modality mod2 should equal input file"
-    )
-
-    assert s1.n_vars == input_h5mu.n_vars, (
-        "number of variables of split file s1 should equal input file"
-    )
-    assert s2.n_vars == input_h5mu.n_vars, (
-        "number of variables of split file s1 should equal input file"
-    )
-
-    assert s1.mod["mod1"].n_vars == input_h5mu.mod["mod1"].n_vars, (
-        "number of variables of split file s1 modalitty mod1 should equal input file"
-    )
-    assert s1.mod["mod2"].n_vars == input_h5mu.mod["mod1"].n_vars, (
-        "number of variables of split file s1 modalitty mod2 should equal input file"
-    )
-
-    assert s2.mod["mod1"].n_vars == input_h5mu.mod["mod1"].n_vars, (
-        "number of variables of split file s2 modalitty mod1 should equal input file"
-    )
-    assert s2.mod["mod2"].n_vars == input_h5mu.mod["mod1"].n_vars, (
-        "number of variables of split file s2 modalitty mod2 should equal input file"
-    )
-
-    # check correct sample splitting
-    assert np.all(s1.mod["mod1"].obs["Obs"] == "A"), (
-        "observation of .obs Obs in s1 should equal A"
-    )
-    assert np.all(s2.mod["mod1"].obs["Obs"] == "B"), (
-        "observation of .obs Obs in s2 should equal B"
-    )
-
-    # Check contents of csv file
-    expected_csv_output = dedent(
-        f"""\
-        name,filename
-        A,{s1_file.name}
-        B,{s2_file.name}
-        """
-    )
-    with open(output_files, "r") as open_csv_file:
-        result = open_csv_file.read()
-        assert result == expected_csv_output
-
-
-def test_sample_split_dropna(run_component, random_path, input_h5mu, input_h5mu_path):
-    output_dir = random_path()
-    output_files = random_path(extension="csv")
-    args = [
-        "--input",
-        input_h5mu_path,
-        "--output",
-        str(output_dir),
-        "--modality",
-        "mod1",
-        "--obs_feature",
-        "Obs",
-        "--drop_obs_nan",
-        "true",
-        "--output_files",
-        str(output_files),
-    ]
-
-    run_component(args)
-
-    # check output dir and file names
-    s1_file = output_dir / f"{input_h5mu_path.stem}_A.h5mu"
-    s2_file = output_dir / f"{input_h5mu_path.stem}_B.h5mu"
-
-    # check that .obs columns with only nan values are dropped correctly
-    s1 = mu.read_h5mu(s1_file)
-    s2 = mu.read_h5mu(s2_file)
-
-    assert s1.n_obs == input_h5mu.n_obs, (
-        "number of observations of split file does not match input file"
-    )
-    assert s2.n_obs == input_h5mu.n_obs, (
-        "number of observations of split file does not match input file"
-    )
-
-    assert s1.mod["mod1"].n_obs == 1, (
-        "number of observations of split file s1 modality mod1 should equal 1"
-    )
-    assert s1.mod["mod2"].n_obs == input_h5mu.n_obs, (
-        "number of observations of split file s1 modality mod2 should equal input file"
-    )
-
-    assert len(s1.mod["mod1"].obs.keys()) == 1, (
-        "number of observation keys split file s1 modality mod1 should equal 1"
-    )
-    assert len(s1.mod["mod2"].obs.keys()) == 1, (
-        "number of observation keys split file s1 modality mod2 should equal 1"
-    )
-
-
-def test_sanitizing(run_component, random_path, input_h5mu_path_non_unique_filenames):
-    output_dir = random_path()
-    output_files = random_path(extension="csv")
-
-    args = [
-        "--input",
-        input_h5mu_path_non_unique_filenames,
-        "--output",
-        str(output_dir),
-        "--modality",
-        "mod3",
-        "--obs_feature",
-        "Obs",
-        "--drop_obs_nan",
-        "true",
-        "--output_files",
-        str(output_files),
-    ]
-
-    with pytest.raises(subprocess.CalledProcessError) as err:
-        run_component(args)
-    assert re.search(
-        r"ValueError: File names are not unique after sanitizing the --obs_feature Obs values",
-        err.value.stdout.decode("utf-8"),
-    )
-
-    args_san = [
-        "--input",
-        input_h5mu_path_non_unique_filenames,
-        "--output",
-        str(output_dir),
-        "--modality",
-        "mod3",
-        "--obs_feature",
-        "Obs",
-        "--drop_obs_nan",
-        "true",
-        "--output_files",
-        str(output_files),
-        "--ensure_unique_filenames",
-        "true",
-    ]
-
-    run_component(args_san)
-
-    # check output dir and file names
-    dir_content = [
-        h5mu_file
-        for h5mu_file in output_dir.iterdir()
-        if h5mu_file.suffix == ".h5mu"
-        and h5mu_file != input_h5mu_path_non_unique_filenames
-    ]
-    s1_file = output_dir / f"{input_h5mu_path_non_unique_filenames.stem}_C_C.h5mu"
-    s2_file = output_dir / f"{input_h5mu_path_non_unique_filenames.stem}_C_C_1.h5mu"
-
-    assert s1_file.is_file(), f"{s1_file} does not exist"
-    assert s2_file.is_file(), f"{s2_file} does not exist"
-    assert set(dir_content) == set([s1_file, s2_file]), (
-        "Output files do not match file names in csv"
-    )
-
-
-if __name__ == "__main__":
-    sys.exit(pytest.main([__file__]))
--- a/src/feature_annotation/spatial_autocorr/config.vsh.yaml
+++ b/src/feature_annotation/spatial_autocorr/config.vsh.yaml
@@ -0,0 +1,111 @@
+name: spatial_autocorr
+namespace: feature_annotation
+label: Spatial Autocorrelation
+summary: Calculate spatial autocorrelation for genes using Moran's I or Geary's C.
+description: |
+  Calculate spatial autocorrelation for genes using Moran's I or Geary's C.
+  This allows identifying spatially variable genes.
+
+argument_groups:
+  - name: Inputs
+    arguments:
+      - name: --input
+        type: file
+        required: true
+        description: Input h5mu file.
+      - name: --modality
+        type: string
+        default: rna
+        description: Modality to use.
+      - name: --layer
+        type: string
+        description: |
+            Layer in the AnnData object to use.
+            - If `attr` is 'X', this is the key in `.layers` to use. If not provided, `.X` is used.
+            - If `attr` is 'obsm', this is the key in `.obsm` to use.
+      - name: --input_genes
+        type: string
+        multiple: true
+        multiple_sep: ","
+        description: |
+          The features to calculate autocorrelation for. Meaning depends on `--attr`:
+          - If `attr` is 'X' (default), this must be a list of gene names from `.var_names`. 
+            If not provided, it defaults to using genes in `.var['highly_variable']` (if present), or all genes.
+            This behavior can be overridden by setting `--use_all_genes` to `true`.
+            Note: You cannot pass a column name from `.var` here.
+          - If `attr` is 'obs', this must be a list of column names from `.obs`. 
+          - If `attr` is 'obsm', this must be indices in `.obsm[layer]` (as strings).
+      - name: --obsp_neighborhood_graph
+        type: string
+        default: spatial_connectivities
+        description: Key in .obsp where spatial connectivities are stored.
+      - name: --use_all_genes
+        type: boolean
+        default: false
+        description: |
+          Whether to use all genes even if highly variable genes are present in .var.
+          If set to true, all genes will be used.
+
+  - name: Parameters
+    arguments:
+      - name: --mode
+        type: string
+        default: moran
+        choices: [moran, geary]
+        description: Mode of spatial autocorrelation.
+      - name: --attr
+        type: string
+        default: X
+        choices: [X, obs, obsm]
+        description: |
+          The attribute of the AnnData object to use for calculation.
+          - 'X': Use gene expression data (default).
+          - 'obs': Use cell metadata.
+          - 'obsm': Use multidimensional embeddings.
+      - name: --n_perms
+        type: integer
+        default: 100
+        description: Number of permutations for p-value calculation.
+      - name: --use_raw
+        type: boolean
+        default: false
+        description: Whether to use .raw attribute of AnnData.
+
+  - name: Outputs
+    arguments:
+      - name: --output
+        type: file
+        direction: output
+        required: true
+        description: Output h5mu file with results in .uns.
+
+resources:
+  - type: python_script
+    path: script.py
+
+test_resources:
+  - type: python_script
+    path: test.py
+  - path: /resources_test/xenium/xenium_tiny.qc.neighbors.h5mu
+
+engines:
+  - type: docker
+    image: python:3.13-slim
+    setup:
+      - type: apt
+        packages:
+          - procps
+      - type: python
+        __merge__: [/src/base/requirements/anndata_mudata.yaml, /src/base/requirements/scanpy.yaml, /src/base/requirements/squidpy.yaml]
+    test_setup:
+      - type: python
+        packages:
+          - pytest
+          - viashpy
+  - type: native
+
+runners:
+  - type: executable
+  - type: nextflow
+    directives:
+      label: [midcpu, midmem]
--- a/src/feature_annotation/spatial_autocorr/script.py
+++ b/src/feature_annotation/spatial_autocorr/script.py
@@ -0,0 +1,80 @@
+import mudata as mu
+import squidpy as sq
+
+## VIASH START
+par = {
+    "input": "resources_test/xenium/xenium_tiny_qc_graph.h5mu",
+    "output": "output.h5mu",
+    "modality": "rna",
+    "mode": "moran",
+    "input_genes": None,
+    "attr": "X",
+    "n_perms": 100,
+    "layer": None,
+    "use_raw": False,
+    "obsp_neighborhood_graph": "spatial_connectivities",
+    "use_all_genes": False,
+}
+## VIASH END
+
+
+def main():
+    print("Reading input MuData...", flush=True)
+    mdata = mu.read_h5mu(par["input"])
+    adata = mdata.mod[par["modality"]]
+
+    # Check for connectivity key
+    if par["obsp_neighborhood_graph"] not in adata.obsp:
+        raise ValueError(
+            f"Connectivity key '{par['obsp_neighborhood_graph']}' not found in .obsp of modality '{par['modality']}'."
+        )
+
+    genes = par["input_genes"]
+    if genes and len(genes) == 0:
+        genes = None
+
+    if genes is None and par["use_all_genes"] and par["attr"] == "X":
+        genes = list(adata.var_names)
+
+    # Handle layer
+    layer = par["layer"]
+    if layer == "None":
+        layer = None
+
+    print(f"Calculating spatial autocorrelation ({par['mode']})...", flush=True)
+
+    # Run Squidpy spatial_autocorr
+    # Note: sq.gr.spatial_autocorr modifies adata in-place, adding results to .uns
+    sq.gr.spatial_autocorr(
+        adata,
+        connectivity_key=par["obsp_neighborhood_graph"],
+        genes=genes,
+        mode=par["mode"],
+        attr=par["attr"],
+        n_perms=par["n_perms"],
+        layer=layer,
+        use_raw=par["use_raw"],
+    )
+
+    result_key = f"{par['mode']}I" if par["mode"] == "moran" else f"{par['mode']}C"
+
+    if result_key in adata.uns:
+        # Log top spatially variable genes
+        df = adata.uns[result_key]
+        if not df.empty:
+            sort_col = "I" if par["mode"] == "moran" else "C"
+            print("Top spatially variable genes:", flush=True)
+            print(df.sort_values(by=sort_col, ascending=False).head(), flush=True)
+    else:
+        print(
+            f"Warning: Expected key '{result_key}' not found in .uns after calculation.",
+            flush=True,
+        )
+
+    print("Writing output...", flush=True)
+    mdata.write_h5mu(par["output"])
+    print("Done!", flush=True)
+
+
+if __name__ == "__main__":
+    main()
--- a/src/feature_annotation/spatial_autocorr/test.py
+++ b/src/feature_annotation/spatial_autocorr/test.py
@@ -0,0 +1,220 @@
+import mudata as mu
+import sys
+import pytest
+import pandas as pd
+
+## VIASH START
+meta = {
+    "resources_dir": "resources_test/xenium",
+    "executable": "./calculate_spatial_autocorrelation",
+}
+## VIASH END
+
+
+def test_calculate_spatial_autocorrelation_moran(run_component, tmp_path):
+    input_path = meta["resources_dir"] + "/xenium_tiny.qc.neighbors.h5mu"
+    output_path = tmp_path / "output_moran.h5mu"
+
+    print(f"Running component with Moran's I on {input_path}")
+
+    run_component(
+        [
+            "--input",
+            str(input_path),
+            "--output",
+            str(output_path),
+            "--modality",
+            "rna",
+            "--obsp_neighborhood_graph",
+            "spatial_connectivities",
+            "--mode",
+            "moran",
+            "--n_perms",
+            "10",  # Reduce permutations for speed
+        ]
+    )
+
+    assert output_path.exists(), "Output file not created"
+
+    mdata = mu.read_h5mu(output_path)
+    adata = mdata.mod["rna"]
+
+    assert "moranI" in adata.uns, "moranI key missing from .uns"
+    df = adata.uns["moranI"]
+    assert isinstance(df, pd.DataFrame), "moranI should be a DataFrame"
+    assert not df.empty, "moranI DataFrame is empty"
+
+    # Check essential columns
+    expected_cols = ["I", "pval_norm", "var_norm"]
+    for col in expected_cols:
+        assert col in df.columns, f"Missing column {col} in moranI dataframe"
+
+    # Check values range
+    assert df["I"].max() <= 1.0, "Moran's I > 1 (theoretical max)"
+    assert df["I"].min() >= -1.0, "Moran's I < -1 (theoretical min)"
+
+
+def test_calculate_spatial_autocorrelation_geary(run_component, tmp_path):
+    input_path = meta["resources_dir"] + "/xenium_tiny.qc.neighbors.h5mu"
+    output_path = tmp_path / "output_geary.h5mu"
+
+    mdata = mu.read_h5mu(input_path)
+    genes = list(mdata.mod["rna"].var_names[:5])  # Test with first 5 genes
+    genes_str = ",".join(genes)
+
+    cmd = [
+        "--input",
+        str(input_path),
+        "--output",
+        str(output_path),
+        "--modality",
+        "rna",
+        "--obsp_neighborhood_graph",
+        "spatial_connectivities",
+        "--mode",
+        "geary",
+        "--n_perms",
+        "10",
+        "--input_genes",
+        genes_str,
+    ]
+
+    print(
+        f"Running component with Geary's C on {input_path} for subset of genes: {genes_str}"
+    )
+    run_component(cmd)
+
+    assert output_path.exists()
+
+    mdata = mu.read_h5mu(output_path)
+    adata = mdata.mod["rna"]
+
+    assert "gearyC" in adata.uns
+    df = adata.uns["gearyC"]
+
+    assert len(df) == 5, f"Expected results for 5 genes, got {len(df)}"
+    assert all(g in df.index for g in genes), (
+        "Not all requested genes are in output index"
+    )
+
+    assert "C" in df.columns
+
+
+def test_calculate_spatial_autocorrelation_obs(run_component, tmp_path):
+    input_path = meta["resources_dir"] + "/xenium_tiny.qc.neighbors.h5mu"
+    output_path = tmp_path / "output_obs.h5mu"
+
+    features = "total_counts,cell_area,nucleus_area"
+
+    print(
+        f"Running component with Moran's I on {input_path} for obs features: {features}"
+    )
+
+    run_component(
+        [
+            "--input",
+            str(input_path),
+            "--output",
+            str(output_path),
+            "--modality",
+            "rna",
+            "--obsp_neighborhood_graph",
+            "spatial_connectivities",
+            "--mode",
+            "moran",
+            "--attr",
+            "obs",
+            "--input_genes",
+            features,
+            "--n_perms",
+            "10",
+        ]
+    )
+
+    assert output_path.exists()
+
+    mdata = mu.read_h5mu(output_path)
+    adata = mdata.mod["rna"]
+
+    assert "moranI" in adata.uns
+    df = adata.uns["moranI"]
+
+    assert len(df) == 3, f"Expected results for 3 features, got {len(df)}"
+    # When attr='obs', the index contains the column names
+    for feature in features.split(","):
+        assert feature in df.index, f"Feature {feature} missing from index: {df.index}"
+
+
+def test_calculate_spatial_autocorrelation_use_all_genes(run_component, tmp_path):
+    input_path = meta["resources_dir"] + "/xenium_tiny.qc.neighbors.h5mu"
+    temp_input_path = tmp_path / "xenium_tiny_hvg.h5mu"
+    output_path_hvg = tmp_path / "output_hvg.h5mu"
+    output_path_all = tmp_path / "output_all.h5mu"
+
+    # Create dataset with highly variable genes
+    mdata = mu.read_h5mu(input_path)
+    adata = mdata.mod["rna"]
+
+    # Mark first 5 genes as highly variable
+    adata.var["highly_variable"] = False
+    genes = list(adata.var_names)
+    for g in genes[:5]:
+        adata.var.loc[g, "highly_variable"] = True
+
+    mdata.write_h5mu(temp_input_path)
+
+    # Test 1: Default behavior (should use only HVG)
+    print("Running component with default settings (expecting HVG usage)")
+    run_component(
+        [
+            "--input",
+            str(temp_input_path),
+            "--output",
+            str(output_path_hvg),
+            "--modality",
+            "rna",
+            "--obsp_neighborhood_graph",
+            "spatial_connectivities",
+            "--mode",
+            "moran",
+            "--n_perms",
+            "10",
+        ]
+    )
+
+    assert output_path_hvg.exists()
+    mdata_hvg = mu.read_h5mu(output_path_hvg)
+    df_hvg = mdata_hvg.mod["rna"].uns["moranI"]
+    assert len(df_hvg) == 5, f"Expected 5 genes (HVG), got {len(df_hvg)}"
+
+    # Test 2: Override to use all genes
+    print("Running component with --use_all_genes (expecting all genes)")
+    run_component(
+        [
+            "--input",
+            str(temp_input_path),
+            "--output",
+            str(output_path_all),
+            "--modality",
+            "rna",
+            "--obsp_neighborhood_graph",
+            "spatial_connectivities",
+            "--mode",
+            "moran",
+            "--n_perms",
+            "10",
+            "--use_all_genes",
+            "true",
+        ]
+    )
+
+    assert output_path_all.exists()
+    mdata_all = mu.read_h5mu(output_path_all)
+    df_all = mdata_all.mod["rna"].uns["moranI"]
+    assert len(df_all) == len(genes), (
+        f"Expected all {len(genes)} genes, got {len(df_all)}"
+    )
+
+
+if __name__ == "__main__":
+    sys.exit(pytest.main([__file__]))
--- a/src/feature_annotation/xenium_spatial_statistics/config.vsh.yaml
+++ b/src/feature_annotation/xenium_spatial_statistics/config.vsh.yaml
@@ -0,0 +1,109 @@
+name: xenium_spatial_statistics
+namespace: feature_annotation
+label: Xenium Spatial Statistics
+description: |
+  Calculate various spatial statistics from Xenium data including cell morphology
+  ratios, position-based features, local density metrics, and global spatial patterns.
+  This component expects a MuData object with a pre-computed spatial neighborhood graph
+  in `.obsp`.
+
+argument_groups:
+  - name: Inputs
+    arguments:
+      - name: --input
+        type: file
+        required: true
+        example: input.h5mu
+        description: |
+          A MuData file containing spatial transcriptomics data and a
+          pre-computed spatial neighborhood graph in `.obsp`.
+
+      - name: --modality
+        type: string
+        required: true
+        default: "rna"
+        description: |
+          The name of the modality to use from the MuData object.
+          This specifies which AnnData object contains the spatial data.
+
+      - name: --obsm_spatial_coordinates
+        type: string
+        default: "spatial"
+        description: |
+          The key in `.obsm` where spatial coordinates are stored.
+          Expected shape is (n_cells, 2) for 2D coordinates.
+
+  - name: Outputs
+    arguments:
+      - name: --output
+        type: file
+        direction: output
+        default: output.h5mu
+        description: |
+          The output MuData file with calculated spatial statistics.
+          Per-cell metrics are stored in `.obs` and global statistics in `.uns`.
+
+      - name: --output_prefix
+        type: string
+        default: "spatial_"
+        description: |
+          Prefix to add to all generated column names in `.obs`.
+
+  - name: Parameters
+    arguments:
+      - name: --density_bandwidth
+        type: double
+        default: 50.0
+        description: |
+          Bandwidth parameter for Gaussian kernel density estimation in spatial units
+          (typically microns). Larger values create smoother density estimates.
+
+      - name: "--calculate_ripley_l"
+        type: boolean
+        default: false
+        description: |
+          Whether to calculate Ripley's L statistic. 
+          Warning: This is an O(N^2) operation and can be very slow for large datasets (>5000 cells).
+
+      - name: "--n_subsample_ripley"
+        type: integer
+        default: -1
+        description: |
+          Number of cells to subsample for Ripley's L calculation. 
+          If -1, use all cells. Recommended to keep this below 5000 for performance.
+
+resources:
+  - type: python_script
+    path: script.py
+
+test_resources:
+  - type: python_script
+    path: test.py
+  - path: /resources_test/xenium/xenium_tiny.qc.neighbors.h5mu
+
+engines:
+  - type: docker
+    image: python:3.13-slim
+    setup:
+      - type: apt
+        packages:
+          - procps
+      - type: python
+        __merge__: 
+          - /src/base/requirements/anndata_mudata.yaml
+          - /src/base/requirements/scanpy.yaml
+          - /src/base/requirements/squidpy.yaml
+        packages:
+          - scikit-learn
+    test_setup:
+      - type: python
+        packages:
+          - pytest
+          - viashpy
+  - type: native
+
+runners:
+  - type: executable
+  - type: nextflow
+    directives:
+      label: [singlecpu, midmem, lowdisk]
--- a/src/feature_annotation/xenium_spatial_statistics/script.py
+++ b/src/feature_annotation/xenium_spatial_statistics/script.py
@@ -0,0 +1,315 @@
+import sys
+import warnings
+
+import mudata as mu
+import numpy as np
+import squidpy as sq
+from scipy.spatial import ConvexHull, Voronoi, distance_matrix
+from sklearn.neighbors import KernelDensity
+
+## VIASH START
+par = {
+    "input": "resources_test/xenium/xenium_tiny_qc.h5mu",
+    "output": "output.h5mu",
+    "modality": "rna",
+    "obsm_spatial_coordinates": "spatial",
+    "density_bandwidth": 50.0,
+    "calculate_ripley_l": False,
+    "n_subsample_ripley": -1,
+    "output_prefix": "spatial_",
+}
+## VIASH END
+
+
+def calculate_morphology_metrics(adata, prefix):
+    """Calculate cell morphology ratios and features."""
+    print("  Calculating morphology metrics...", flush=True)
+
+    if "cell_area" not in adata.obs or "nucleus_area" not in adata.obs:
+        warnings.warn(
+            "cell_area and/or nucleus_area not found in .obs. "
+            "Skipping morphology metrics."
+        )
+        return
+
+    # Nucleus to cell area ratio - can be removed once https://github.com/openpipelines-bio/openpipeline_qc/issues/18 is fixed.
+    adata.obs[f"{prefix}nucleus_cell_ratio"] = (
+        adata.obs["nucleus_area"] / adata.obs["cell_area"]
+    )
+
+    # Cell area percentile (relative size)
+    adata.obs[f"{prefix}cell_area_percentile"] = (
+        adata.obs["cell_area"].rank(pct=True) * 100
+    )
+
+    print(
+        f"    Added: {prefix}nucleus_cell_ratio, {prefix}cell_area_percentile",
+        flush=True,
+    )
+
+
+def calculate_position_features(adata, spatial_coords, prefix):
+    """Calculate position-based features."""
+    print("  Calculating position-based features...", flush=True)
+
+    # Tissue centroid
+    centroid = spatial_coords.mean(axis=0)
+
+    # Distance to centroid
+    distances_to_centroid = np.linalg.norm(spatial_coords - centroid, axis=1)
+    adata.obs[f"{prefix}distance_to_centroid"] = distances_to_centroid
+
+    # Normalized coordinates (0-1 scale)
+    min_coords = spatial_coords.min(axis=0)
+    max_coords = spatial_coords.max(axis=0)
+    coord_range = max_coords - min_coords
+
+    normalized_coords = (spatial_coords - min_coords) / coord_range
+    adata.obs[f"{prefix}norm_x"] = normalized_coords[:, 0]
+    adata.obs[f"{prefix}norm_y"] = normalized_coords[:, 1]
+
+    # Distance to convex hull boundary
+    try:
+        hull = ConvexHull(spatial_coords)
+        hull_points = spatial_coords[hull.vertices]
+
+        # For each point, find distance to nearest hull vertex (approximation)
+        distances_to_boundary = np.min(
+            distance_matrix(spatial_coords, hull_points), axis=1
+        )
+        adata.obs[f"{prefix}distance_to_boundary"] = distances_to_boundary
+        print(
+            "    Added: distance_to_centroid, norm_x, norm_y, distance_to_boundary",
+            flush=True,
+        )
+    except Exception as e:
+        warnings.warn(f"Could not calculate convex hull: {e}")
+        print("    Added: distance_to_centroid, norm_x, norm_y", flush=True)
+
+
+def calculate_density_metrics(adata, spatial_coords, bandwidth, prefix):
+    """Calculate local density metrics."""
+    print("  Calculating density metrics...", flush=True)
+
+    # Kernel density estimation
+    kde = KernelDensity(bandwidth=bandwidth, kernel="gaussian")
+    kde.fit(spatial_coords)
+    log_density = kde.score_samples(spatial_coords)
+    adata.obs[f"{prefix}kernel_density"] = np.exp(log_density)
+
+    # Calculate degree centrality (as a proxy for local density / number of neighbors)
+    # This assumes spatial_connectivities is already present in .obsp (from build_spatial_graph)
+    if "spatial_connectivities" in adata.obsp:
+        spatial_connectivities = adata.obsp["spatial_connectivities"]
+        degrees = spatial_connectivities.sum(axis=1)
+        # If sparse matrix, it returns matrix object, need to convert to array
+        if hasattr(degrees, "A1"):
+            degrees = degrees.A1
+        adata.obs[f"{prefix}graph_degree"] = degrees
+
+        print("    Added: kernel_density, graph_degree", flush=True)
+    else:
+        print(
+            "    Warning: 'spatial_connectivities' not found in .obsp. Skipping graph_degree.",
+            flush=True,
+        )
+        print("    Added: kernel_density", flush=True)
+
+
+def calculate_voronoi_metrics(adata, spatial_coords, prefix):
+    """Calculate Voronoi tessellation statistics."""
+    print("  Calculating Voronoi tessellation...", flush=True)
+
+    try:
+        vor = Voronoi(spatial_coords)
+
+        polygon_areas = []
+        neighbor_counts = []
+
+        for point_idx in range(len(spatial_coords)):
+            region_idx = vor.point_region[point_idx]
+            region = vor.regions[region_idx]
+
+            # Skip infinite regions
+            if -1 in region or len(region) == 0:
+                polygon_areas.append(np.nan)
+                neighbor_counts.append(np.nan)
+                continue
+
+            # Calculate polygon area using shoelace formula
+            vertices = vor.vertices[region]
+            x = vertices[:, 0]
+            y = vertices[:, 1]
+            area = 0.5 * np.abs(np.dot(x, np.roll(y, 1)) - np.dot(y, np.roll(x, 1)))
+            polygon_areas.append(area)
+
+            # Count neighbors (shared vertices)
+            neighbor_counts.append(len(region))
+
+        adata.obs[f"{prefix}voronoi_area"] = polygon_areas
+        adata.obs[f"{prefix}voronoi_neighbors"] = neighbor_counts
+
+        print("    Added: voronoi_area, voronoi_neighbors", flush=True)
+    except Exception as e:
+        warnings.warn(f"Could not calculate Voronoi tessellation: {e}")
+
+
+def calculate_global_statistics(adata, spatial_coords, par):
+    """Calculate global spatial pattern statistics."""
+    print("  Calculating global spatial statistics...", flush=True)
+
+    stats = {}
+
+    # Global density
+    try:
+        hull = ConvexHull(spatial_coords)
+        area = hull.volume  # In 2D, volume of convex hull is the area
+        stats["area_calculation_method"] = "convex_hull"
+        stats["cell_density"] = len(spatial_coords) / area
+        stats["total_area"] = area
+        stats["n_cells"] = len(spatial_coords)
+    except Exception as e:
+        print(f"    Error: Could not calculate convex hull area ({e})", flush=True)
+
+    # Ripley's L (using squidpy)
+    if par["calculate_ripley_l"]:
+        print("    Computing Ripley's L function (Squidpy)...", flush=True)
+
+        n_subsample = par["n_subsample_ripley"]
+
+        try:
+            # Create a working AnnData for Ripley's
+            if n_subsample > 0 and len(adata) > n_subsample:
+                print(f"      Subsampling to {n_subsample} random cells...", flush=True)
+
+                # Use numpy to generate random indices
+                indices = np.random.choice(len(adata), n_subsample, replace=False)
+                adata_ripley = adata[indices].copy()
+                stats["ripley_l_subsampled"] = True
+            else:
+                adata_ripley = adata.copy()
+                stats["ripley_l_subsampled"] = False
+
+            # Squidpy requires a cluster key. We create a dummy one for "global" context.
+            adata_ripley.obs["_temp_global"] = "all"
+            adata_ripley.obs["_temp_global"] = adata_ripley.obs["_temp_global"].astype(
+                "category"
+            )
+
+            # Calculate Ripley's L statistic
+            # This stores the result in adata.uns['all_L']
+            sq.gr.ripley(
+                adata_ripley,
+                cluster_key="_temp_global",
+                mode="L",
+                spatial_key=par["obsm_spatial_coordinates"],
+                n_simulations=50,
+            )
+
+            # Extract basic stats from the results
+            if "all_L" in adata_ripley.uns and "L_stat" in adata_ripley.uns["all_L"]:
+                l_stat_df = adata_ripley.uns["all_L"]["L_stat"]
+                stats["ripley_l_max"] = float(l_stat_df.max().max())
+                stats["ripley_l_mean"] = float(l_stat_df.mean().mean())
+
+            # Clean up
+            if "all_L" in adata_ripley.uns:
+                del adata_ripley.uns["all_L"]
+
+        except Exception as e:
+            print(f"    Error calculating Ripley's L: {e}", flush=True)
+            import traceback
+
+            traceback.print_exc()
+    else:
+        print("    Skipping Ripley's L (disabled in config)...", flush=True)
+
+    # Spatial extent
+    min_coords = spatial_coords.min(axis=0)
+    max_coords = spatial_coords.max(axis=0)
+
+    stats["spatial_extent_x"] = float(max_coords[0] - min_coords[0])
+    stats["spatial_extent_y"] = float(max_coords[1] - min_coords[1])
+    stats["centroid_x"] = float(spatial_coords[:, 0].mean())
+    stats["centroid_y"] = float(spatial_coords[:, 1].mean())
+
+    if "cell_density" in stats:
+        print(
+            f"    Global stats: cell_density={stats['cell_density']:.4f}",
+            flush=True,
+        )
+
+    return stats
+
+
+def main(par):
+    print(f"\n>>> Reading MuData from '{par['input']}'...", flush=True)
+    mdata = mu.read_h5mu(par["input"])
+    print(mdata, flush=True)
+
+    print(f"\n>>> Extracting modality '{par['modality']}'...", flush=True)
+    if par["modality"] not in mdata.mod:
+        raise KeyError(
+            f"Modality '{par['modality']}' not found in MuData. "
+            f"Available modalities: {list(mdata.mod.keys())}"
+        )
+    adata = mdata[par["modality"]]
+    print(adata, flush=True)
+
+    print(
+        f"\n>>> Extracting spatial coordinates from .obsm['{par['obsm_spatial_coordinates']}']...",
+        flush=True,
+    )
+    if par["obsm_spatial_coordinates"] not in adata.obsm:
+        raise KeyError(
+            f"Spatial key '{par['obsm_spatial_coordinates']}' not found in .obsm. "
+            f"Available keys: {list(adata.obsm.keys())}"
+        )
+
+    spatial_coords = adata.obsm[par["obsm_spatial_coordinates"]]
+    if spatial_coords.shape[1] != 2:
+        raise ValueError(
+            f"Expected 2D spatial coordinates, got shape {spatial_coords.shape}"
+        )
+    print(f"  Shape: {spatial_coords.shape} (n_cells × 2)", flush=True)
+
+    prefix = par["output_prefix"]
+
+    # Calculate morphology metrics
+    print("\n>>> Calculating morphology metrics...", flush=True)
+    calculate_morphology_metrics(adata, prefix)
+
+    # Calculate position features
+    print("\n>>> Calculating position-based features...", flush=True)
+    calculate_position_features(adata, spatial_coords, prefix)
+
+    # Calculate density metrics
+    print("\n>>> Calculating density metrics...", flush=True)
+    calculate_density_metrics(
+        adata,
+        spatial_coords,
+        par["density_bandwidth"],
+        prefix,
+    )
+
+    # Calculate Voronoi tessellation
+    print("\n>>> Calculating Voronoi tessellation...", flush=True)
+    calculate_voronoi_metrics(adata, spatial_coords, prefix)
+
+    # Calculate global statistics
+    print("\n>>> Calculating global spatial statistics...", flush=True)
+    global_stats = calculate_global_statistics(adata, spatial_coords, par)
+
+    # Store in uns
+    if "spatial_stats" not in adata.uns:
+        adata.uns["spatial_stats"] = {}
+    adata.uns["spatial_stats"].update(global_stats)
+
+    print(f"\n>>> Writing output to '{par['output']}'...", flush=True)
+    mdata.write_h5mu(par["output"])
+
+    print("\n>>> Done!\n", flush=True)
+
+
+if __name__ == "__main__":
+    sys.exit(main(par))
--- a/src/feature_annotation/xenium_spatial_statistics/test.py
+++ b/src/feature_annotation/xenium_spatial_statistics/test.py
@@ -0,0 +1,75 @@
+import mudata as mu
+import sys
+import pytest
+
+## VIASH START
+meta = {
+    "resources_dir": "resources_test/xenium",
+    "executable": "./xenium_spatial_statistics",
+}
+## VIASH END
+
+
+def test_calculate_spatial_statistics(run_component, tmp_path):
+    input_path = meta["resources_dir"] + "/xenium_tiny.qc.neighbors.h5mu"
+    output_path = tmp_path / "output.h5mu"
+
+    # Run the component directly on the input file
+    run_component(
+        [
+            "--input",
+            str(input_path),
+            "--output",
+            str(output_path),
+            "--modality",
+            "rna",
+            "--obsm_spatial_coordinates",
+            "spatial",
+            "--density_bandwidth",
+            "50.0",
+            "--calculate_ripley_l",
+            "false",
+        ]
+    )
+
+    assert output_path.exists(), "Output file not created"
+
+    mdata_out = mu.read_h5mu(output_path)
+    adata_out = mdata_out.mod["rna"]
+
+    # Check for expected columns in .obs
+    expected_cols = [
+        "spatial_nucleus_cell_ratio",
+        "spatial_cell_area_percentile",
+        "spatial_distance_to_centroid",
+        "spatial_norm_x",
+        "spatial_norm_y",
+        "spatial_kernel_density",
+        "spatial_graph_degree",
+        "spatial_voronoi_area",
+        "spatial_voronoi_neighbors",
+    ]
+
+    for col in expected_cols:
+        assert col in adata_out.obs.columns, f"Column {col} missing from output .obs"
+
+    # Check for expected metrics in .uns['spatial_stats']
+    assert "spatial_stats" in adata_out.uns, "spatial_stats missing from .uns"
+    stats = adata_out.uns["spatial_stats"]
+
+    expected_stats = [
+        "cell_density",
+        "total_area",
+        "n_cells",
+        "spatial_extent_x",
+        "spatial_extent_y",
+        "centroid_x",
+        "centroid_y",
+    ]
+
+    for stat in expected_stats:
+        assert stat in stats, f"Stat {stat} missing from spatial_stats"
+
+
+if __name__ == "__main__":
+    sys.exit(pytest.main([__file__]))
--- a/src/neighbors/join_graphs/config.vsh.yaml
+++ b/src/neighbors/join_graphs/config.vsh.yaml
@@ -0,0 +1,91 @@
+name: join_graphs
+namespace: neighbors
+description: |
+  Combine spatial and expression neighborhood graphs into a single graph for 
+  use in downstream clustering or trajectory analysis.
+  
+  The fusion is a linear combination of the expression connectivities 
+  and spatial connectivities matrices, weighted by `alpha`.
+  
+  $C_{joint} = (1 - \alpha) \cdot C_{expression} + \alpha \cdot C_{spatial}$
+
+authors:
+  - __merge__: /src/authors/jakub_majercik.yaml
+    roles: [ maintainer ]
+
+argument_groups:
+  - name: "Inputs"
+    arguments:
+      - name: "--input"
+        type: file
+        required: true
+        description: Input H5MU file with pre-computed expression and spatial neighborhood graphs.
+      - name: "--modality"
+        type: string
+        default: "rna"
+        required: false
+        description: Modality from the input MuData file to process.
+      - name: "--input_obsp_expression_graph"
+        type: string
+        default: "connectivities"
+        description: |
+          Key in `adata.obsp` containing the expression connectivity matrix.
+      - name: "--input_obsp_spatial_graph"
+        type: string
+        default: "spatial_connectivities"
+        description: |
+          Key in `adata.obsp` containing the spatial connectivity matrix.
+
+  - name: "Parameters"
+    arguments:
+      - name: "--alpha"
+        type: double
+        default: 0.5
+        min: 0.0
+        max: 1.0
+        description: |
+          Weight of the spatial graph in the connection fusion. 
+          A value of 0 results in the original expression graph; 
+          a value of 1 results in the spatial graph only.
+
+  - name: Outputs
+    arguments:
+      - name: "--output"
+        type: file
+        direction: output
+        required: true
+        description: Output H5MU file path.
+        example: output.h5mu
+      - name: "--output_obsp_graph"
+        type: string
+        default: "spatial_expression_connectivities"
+        description: |
+          Key under which to store the fused connectivity matrix in `adata.obsp`.
+    __merge__: [., /src/base/h5_compression_argument.yaml]
+
+resources:
+  - type: python_script
+    path: script.py
+  - path: /src/utils/setup_logger.py
+
+test_resources:
+  - type: python_script
+    path: test.py
+  - path: /resources_test/xenium/xenium_tiny.qc.all_neighbors.pca.h5mu
+
+engines:
+  - type: docker
+    image: python:3.12-slim
+    setup:
+      - type: apt
+        packages:
+          - procps
+      - type: python
+        __merge__: [ /src/base/requirements/scanpy.yaml, /src/base/requirements/anndata_mudata.yaml, . ]
+    __merge__: [ /src/base/requirements/python_test_setup.yaml, .]
+
+runners:
+  - type: executable
+  - type: nextflow
+    directives:
+      label: [midcpu, midmem, middisk]
--- a/src/neighbors/join_graphs/script.py
+++ b/src/neighbors/join_graphs/script.py
@@ -0,0 +1,63 @@
+import sys
+import mudata as mu
+
+## VIASH START
+par = {
+    # Inputs
+    "input": "resources_test/xenium/xenium_tiny.spatial_expression_neighbors.h5mu",
+    "modality": "rna",
+    "input_obsp_expression_graph": "connectivities",
+    "input_obsp_spatial_graph": "spatial_connectivities",
+    # Fusion options
+    "alpha": 0.2,
+    # Outputs
+    "output": "foo.h5mu",
+    "output_compression": None,
+    "output_obsp_graph": "spatial_expression_connectivities",
+}
+meta = {"resources_dir": "src/utils/"}
+## VIASH END
+
+sys.path.append(meta["resources_dir"])
+from setup_logger import setup_logger
+
+logger = setup_logger()
+
+## Read data
+logger.info("Reading input data...")
+adata = mu.read_h5ad(par["input"], mod=par["modality"])
+
+## Validate inputs
+spatial_key = par["input_obsp_spatial_graph"]
+if spatial_key not in adata.obsp:
+    raise ValueError(f"Spatial graph key '{spatial_key}' not found in .obsp.")
+
+expr_key = par["input_obsp_expression_graph"]
+if expr_key not in adata.obsp:
+    raise ValueError(f"Expression graph key '{expr_key}' not found in .obsp.")
+
+nn_graph_genes = adata.obsp[expr_key]
+nn_graph_space = adata.obsp[spatial_key]
+
+## Combine graphs
+alpha = par["alpha"]
+logger.info(
+    f"Combining graphs (alpha={alpha}: spatial weight, "
+    f"{1 - alpha:.2f}: expression weight)..."
+)
+joint_graph = (1 - alpha) * nn_graph_genes + alpha * nn_graph_space
+out_key = par["output_obsp_graph"]
+logger.info(f"Storing result in .obsp['{out_key}']...")
+adata.obsp[out_key] = joint_graph
+adata.uns[out_key] = {
+    "params": {"alpha": alpha},
+    "inputs": {
+        "expression_graph": expr_key,
+        "spatial_graph": spatial_key,
+    },
+}
+
+## Write output
+logger.info("Saving output data...")
+mdata = mu.MuData({par["modality"]: adata})
+mdata.write_h5mu(par["output"], compression=par["output_compression"])
--- a/src/neighbors/join_graphs/test.py
+++ b/src/neighbors/join_graphs/test.py
@@ -0,0 +1,103 @@
+import pytest
+import mudata as mu
+import sys
+
+## VIASH START
+meta = {
+    "executable": "./target/executable/neighbors/join_graphs/join_graphs",
+    "resources_dir": "resources_test/xenium/",
+}
+## VIASH END
+
+input_file = f"{meta['resources_dir']}/xenium_tiny.qc.all_neighbors.pca.h5mu"
+
+
+def test_default_execution(run_component, tmp_path):
+    output = tmp_path / "fused_neighbors.h5mu"
+
+    run_component(
+        [
+            "--input",
+            input_file,
+            "--output",
+            str(output),
+            "--output_obsp_graph",
+            "fused_conn",
+            "--alpha",
+            "0.2",
+            "--output_compression",
+            "gzip",
+        ]
+    )
+
+    assert output.is_file(), "Output file was not created."
+    mdata = mu.read_h5mu(output)
+    assert "rna" in mdata.mod, "Expected 'rna' modality in output."
+    adata = mdata.mod["rna"]
+
+    assert "fused_conn" in adata.obsp.keys(), (
+        "Expected output obsp connectivities key to be present."
+    )
+    assert "fused_conn" in adata.uns.keys(), "Expected metadata for fusion in .uns."
+    assert adata.uns["fused_conn"]["params"]["alpha"] == 0.2, (
+        "Expected alpha parameter to be stored in .uns."
+    )
+
+    # Check graph properties (should be combination of input graphs)
+    fused = adata.obsp["fused_conn"]
+    spatial = adata.obsp["spatial_connectivities"]
+    expr = adata.obsp["connectivities"]
+
+    assert fused.shape == expr.shape == spatial.shape, (
+        "Fused graph should have same shape as input graphs."
+    )
+
+
+def test_alpha_zero_equals_expression(run_component, tmp_path):
+    output = tmp_path / "alpha_zero.h5mu"
+    run_component(
+        [
+            "--input",
+            input_file,
+            "--output",
+            str(output),
+            "--alpha",
+            "0.0",
+            "--output_obsp_graph",
+            "fused_zero",
+        ]
+    )
+    mdata = mu.read_h5mu(output)
+    adata = mdata.mod["rna"]
+
+    fused = adata.obsp["fused_zero"]
+    # Should equal expression graph exactly (dense check might be heavy, check nnz or sum)
+    expr = adata.obsp["connectivities"]
+    assert abs(fused.sum() - expr.sum()) < 1e-6
+
+
+def test_alpha_one_equals_spatial(run_component, tmp_path):
+    output = tmp_path / "alpha_one.h5mu"
+    run_component(
+        [
+            "--input",
+            input_file,
+            "--output",
+            str(output),
+            "--alpha",
+            "1.0",
+            "--output_obsp_graph",
+            "fused_one",
+        ]
+    )
+    mdata = mu.read_h5mu(output)
+    adata = mdata.mod["rna"]
+
+    fused = adata.obsp["fused_one"]
+    # Should equal spatial graph exactly
+    spatial = adata.obsp["spatial_connectivities"]
+    assert abs(fused.sum() - spatial.sum()) < 1e-6
+
+
+if __name__ == "__main__":
+    sys.exit(pytest.main([__file__]))
--- a/src/nichecompass/nichecompass/config.vsh.yaml
+++ b/src/nichecompass/nichecompass/config.vsh.yaml
@@ -33,6 +33,12 @@ argument_groups:
        type: string
        required: false
        description: "Input layer to use. If None, X is used"
+      - name: "--var_input"
+        type: string
+        default: "filter_with_hvg"
+        description: |
+          Key in adata.var that indicates which features to include as input for the NicheCompass model.
+          If not provided, all features will be included.
      - name: "--input_obsp_spatial_connectivities"
        type: string
        default: "spatial_connectivities"
@@ -429,6 +435,7 @@ resources:
  - type: python_script
    path: script.py
  - path: /src/utils/setup_logger.py
+  - path: /src/utils/subset_vars.py

 test_resources:
  - type: python_script
--- a/src/nichecompass/nichecompass/script.py
+++ b/src/nichecompass/nichecompass/script.py
@@ -15,6 +15,7 @@ par = {
    "input_gp_mask": "resources_test/niche/prior_knowledge_gp_mask.json",
    "input_obs_covariates": None,
    "input_obsp_spatial_connectivities": "spatial_connectivities",
+    "var_input": "filter_with_hvg",
    ## GP Mask
    "min_genes_per_gp": 2,
    "min_source_genes_per_gp": 1,
@@ -89,6 +90,7 @@ meta = {"resources_dir": "src/utils/"}

 sys.path.append(meta["resources_dir"])
 from setup_logger import setup_logger
+from subset_vars import subset_vars

 logger = setup_logger()

@@ -101,6 +103,11 @@ logger.info(f"GPU count: {torch.cuda.device_count()}")
 ## Read in data
 adata = mu.read_h5ad(par["input"], mod=par["modality"])

+# Subset to HVG
+if par["var_input"]:
+    # Subset to HVG
+    adata = subset_vars(adata, subset_col=par["var_input"]).copy()
+
 # Counts need to be float32 to be processed by nichecompass model
 # See https://discuss.pytorch.org/t/runtimeerror-mat1-and-mat2-must-have-the-same-dtype/166759
 counts_dtype = (
--- a/src/nichecompass/nichecompass/test.py
+++ b/src/nichecompass/nichecompass/test.py
@@ -1,5 +1,6 @@
 import pytest
 import mudata as mu
+import numpy as np
 import sys

 ## VIASH START
@@ -8,21 +9,38 @@ meta = {
 }
 ## VIASH END

-input_xenium = f"{meta['resources_dir']}/xenium_tiny.h5mu"
-input_cosmx = f"{meta['resources_dir']}/Lung5_Rep2_tiny.h5mu"
+input_path_xenium = f"{meta['resources_dir']}/xenium_tiny.h5mu"
+input_path_cosmx = f"{meta['resources_dir']}/Lung5_Rep2_tiny.h5mu"
 gp_mask = f"{meta['resources_dir']}/prior_knowledge_gp_mask.json"


-def test_simple_execution_xenium(run_component, tmp_path):
+@pytest.fixture
+def input_xenium_with_hvg_filter(tmp_path):
+    mdata = mu.read_h5mu(input_path_xenium)
+    adata = mdata.mod["rna"]
+    np.random.seed(42)
+    n_select = min(1000, adata.n_vars)
+    selected = np.random.choice(adata.n_vars, size=n_select, replace=False)
+    mask = np.zeros(adata.n_vars, dtype=bool)
+    mask[selected] = True
+    adata.var["filter_with_hvg"] = mask
+    output = tmp_path / "xenium_with_hvg_filter.h5mu"
+    mdata.write_h5mu(output)
+    return str(output)
+
+
+def test_simple_execution_xenium(run_component, tmp_path, input_xenium_with_hvg_filter):
    output = tmp_path / "nc_xenium.h5mu"

    # run component
    run_component(
        [
            "--input",
-            input_xenium,
+            input_xenium_with_hvg_filter,
            "--input_gp_mask",
            gp_mask,
+            "--var_input",
+            "filter_with_hvg",
            "--n_epochs",
            "1",
            "--n_epochs_all_gps",
@@ -84,5 +102,48 @@ def test_simple_execution_xenium(run_component, tmp_path):
    ), "Expected GP targets and sources varm to have same shape"


+def test_no_var_input(run_component, tmp_path, input_xenium_with_hvg_filter):
+    output = tmp_path / "nc_xenium_no_var_input.h5mu"
+
+    # run component without var_input (all genes are used)
+    run_component(
+        [
+            "--input",
+            input_xenium_with_hvg_filter,
+            "--input_gp_mask",
+            gp_mask,
+            "--var_input",
+            "",
+            "--n_epochs",
+            "1",
+            "--n_epochs_all_gps",
+            "0",
+            "--n_epochs_no_edge_recon",
+            "0",
+            "--n_epochs_no_cat_covariates_contrastive",
+            "0",
+            "--output",
+            str(output),
+            "--output_model",
+            "test_model_no_var_input",
+            "--output_compression",
+            "gzip",
+        ]
+    )
+
+    assert output.is_file(), "output file was not created"
+    mdata = mu.read_h5mu(output)
+    assert list(mdata.mod.keys()) == ["rna"], "Expected modality rna"
+    adata = mdata.mod["rna"]
+
+    expected_uns_keys = [
+        "nichecompass_gp_names",
+        "nichecompass_active_gp_names",
+    ]
+    assert all([uns in adata.uns.keys() for uns in expected_uns_keys]), (
+        f"Expected uns keys: {expected_uns_keys}, found: {list(adata.uns.keys())}"
+    )
+
+
 if __name__ == "__main__":
    sys.exit(pytest.main([__file__]))
--- a/src/utils/subset_vars.py
+++ b/src/utils/subset_vars.py
@@ -0,0 +1,49 @@
+def subset_vars(adata, subset_col):
+    """
+    Subset AnnData object on highly variable genes or a boolean mask.
+
+    Parameters
+    ----------
+    adata : AnnData
+        Annotated data object
+    subset_col : str, pd.Series, pd.Index, or np.ndarray
+        Name of the boolean column in `adata.var` that contains the information if features should be used or not,
+        or a boolean mask (same length as adata.var)
+
+    Returns
+    -------
+    AnnData
+        Copy of `adata` with subsetted features
+    """
+    import pandas as pd
+    import numpy as np
+
+    # Convert all input types to a pandas Series
+    if isinstance(subset_col, str):
+        if subset_col not in adata.var.columns:
+            raise ValueError(
+                f"Requested to use .var column '{subset_col}' as a selection of genes, but the column is not available."
+            )
+        mask = adata.var[subset_col]
+    elif isinstance(subset_col, pd.Series):
+        mask = subset_col
+    elif isinstance(subset_col, (pd.Index, np.ndarray, list)):
+        mask = pd.Series(subset_col, index=adata.var.index)
+    else:
+        raise TypeError(
+            "subset_col must be a string (column name) or a boolean mask (Series, Index, ndarray, or list)."
+        )
+
+    # Validate mask
+    if not pd.api.types.is_bool_dtype(mask):
+        raise ValueError(
+            f"Expected mask to be boolean, but found {mask.dtype}. Can not subset data."
+        )
+    if mask.isna().sum() > 0:
+        raise ValueError("Mask contains NaN values. Can not subset data.")
+    if len(mask) != adata.n_vars:
+        raise ValueError(
+            f"Mask length {len(mask)} does not match number of variables {adata.n_vars}."
+        )
+
+    return adata[:, mask].copy()
--- a/src/workflows/niche/nichecompass_leiden/config.vsh.yaml
+++ b/src/workflows/niche/nichecompass_leiden/config.vsh.yaml
@@ -57,18 +57,18 @@ argument_groups:
        type: string
        default: "spatial"
        description: "Key in adata.obsm where spatial coordinates are stored"
+      - name: "--var_input"
+        type: string
+        default: "filter_with_hvg"
+        description: |
+          Key in adata.var that indicates which features to include as input for the NicheCompass model.
+          If not provided, all features will be included.

  - name: "Sample ID options"
    description: |
      Options for adding the id to .obs on the MuData object. Having a sample 
      id present in a requirement of several components for this pipeline.
    arguments:
-      - name: "--include_sample_as_covariate"
-        description: |
-          Whether to include the sample information as a categorical covariate for the 
-          NicheCompass model.
-        type: boolean
-        default: true
      - name: "--add_id_to_obs"
        description: "Add the value passed with --id to .obs."
        type: boolean
@@ -357,6 +357,7 @@ dependencies:
  - name: neighbors/spatial_neighborhood_graph
  - name: nichecompass/nichecompass
  - name: dataflow/split_h5mu
+    repository: openpipeline
  - name: workflows/multiomics/neighbors_leiden_umap
    repository: openpipeline

--- a/src/workflows/niche/nichecompass_leiden/main.nf
+++ b/src/workflows/niche/nichecompass_leiden/main.nf
@@ -117,6 +117,7 @@ workflow run_wf {
        "input": state.input,
        "input_gp_mask": state.input_gp_mask,
        "input_obs_covariates": state.input_obs_covariates,
+        "var_input": state.var_input,
        "modality": state.modality,
        "layer": state.layer,
        "min_genes_per_gp": state.min_genes_per_gp,
--- a/src/workflows/niche/nichecompass_leiden/test.nf
+++ b/src/workflows/niche/nichecompass_leiden/test.nf
@@ -19,7 +19,8 @@ workflow test_wf {
        n_epochs_all_gps: 0,
        n_epochs_no_edge_recon: 0,
        n_epochs_no_cat_covariates_contrastive_loss: 0,
-        output_model: "simple_execution_test_model"
+        output_model: "simple_execution_test_model",
+        var_input: null
      ]
    ])
    | map { state -> [state.id, state] }
--- a/target/_private/executable/filter/subset_cosmx/.config.vsh.yaml
+++ b/target/_private/executable/filter/subset_cosmx/.config.vsh.yaml
@@ -112,7 +112,7 @@ repositories:
 - type: "vsh"
  name: "openpipeline"
  repo: "openpipeline"
-  tag: "v4.0.2"
+  tag: "v4.0.3"
 links:
  repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
  docker_registry: "ghcr.io"
@@ -227,7 +227,7 @@ build_info:
  output: "target/_private/executable/filter/subset_cosmx"
  executable: "target/_private/executable/filter/subset_cosmx/subset_cosmx"
  viash_version: "0.9.4"
-  git_commit: "ad19aba349736848d99a9646870d8e7868eb6515"
+  git_commit: "87e62605aafd706d539bf2978ef47ede6fe41926"
  git_remote: "https://github.com/openpipelines-bio/openpipeline_spatial"
 package_config:
  name: "openpipeline_spatial"
@@ -241,7 +241,7 @@ package_config:
  - type: "vsh"
    name: "openpipeline"
    repo: "openpipeline"
-    tag: "v4.0.2"
+    tag: "v4.0.3"
  viash_version: "0.9.4"
  source: "src"
  target: "target"
--- a/target/_private/executable/filter/subset_cosmx/subset_cosmx
+++ b/target/_private/executable/filter/subset_cosmx/subset_cosmx
@@ -458,9 +458,9 @@ RUN pip install --upgrade pip && \

 LABEL org.opencontainers.image.authors="Dorien Roosen, Weiwei Schultz"
 LABEL org.opencontainers.image.description="Companion container for running component filter subset_cosmx"
-LABEL org.opencontainers.image.created="2026-02-17T11:00:57Z"
+LABEL org.opencontainers.image.created="2026-03-25T10:11:21Z"
 LABEL org.opencontainers.image.source="https://github.com/openpipelines-bio/openpipeline_spatial"
-LABEL org.opencontainers.image.revision="ad19aba349736848d99a9646870d8e7868eb6515"
+LABEL org.opencontainers.image.revision="87e62605aafd706d539bf2978ef47ede6fe41926"
 LABEL org.opencontainers.image.version="niche-compass"

 VIASHDOCKER
--- a/target/_private/nextflow/filter/subset_cosmx/.config.vsh.yaml
+++ b/target/_private/nextflow/filter/subset_cosmx/.config.vsh.yaml
@@ -112,7 +112,7 @@ repositories:
 - type: "vsh"
  name: "openpipeline"
  repo: "openpipeline"
-  tag: "v4.0.2"
+  tag: "v4.0.3"
 links:
  repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
  docker_registry: "ghcr.io"
@@ -227,7 +227,7 @@ build_info:
  output: "target/_private/nextflow/filter/subset_cosmx"
  executable: "target/_private/nextflow/filter/subset_cosmx/main.nf"
  viash_version: "0.9.4"
-  git_commit: "ad19aba349736848d99a9646870d8e7868eb6515"
+  git_commit: "87e62605aafd706d539bf2978ef47ede6fe41926"
  git_remote: "https://github.com/openpipelines-bio/openpipeline_spatial"
 package_config:
  name: "openpipeline_spatial"
@@ -241,7 +241,7 @@ package_config:
  - type: "vsh"
    name: "openpipeline"
    repo: "openpipeline"
-    tag: "v4.0.2"
+    tag: "v4.0.3"
  viash_version: "0.9.4"
  source: "src"
  target: "target"
--- a/target/_private/nextflow/filter/subset_cosmx/main.nf
+++ b/target/_private/nextflow/filter/subset_cosmx/main.nf
@@ -3187,7 +3187,7 @@ meta = [
      "type" : "vsh",
      "name" : "openpipeline",
      "repo" : "openpipeline",
-      "tag" : "v4.0.2"
+      "tag" : "v4.0.3"
    }
  ],
  "links" : {
@@ -3333,7 +3333,7 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/_private/nextflow/filter/subset_cosmx",
    "viash_version" : "0.9.4",
-    "git_commit" : "ad19aba349736848d99a9646870d8e7868eb6515",
+    "git_commit" : "87e62605aafd706d539bf2978ef47ede6fe41926",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline_spatial"
  },
  "package_config" : {
@@ -3353,7 +3353,7 @@ meta = [
        "type" : "vsh",
        "name" : "openpipeline",
        "repo" : "openpipeline",
-        "tag" : "v4.0.2"
+        "tag" : "v4.0.3"
      }
    ],
    "viash_version" : "0.9.4",
--- a/target/_test/executable/test_workflows/ingestion/spaceranger_mapping_test/.config.vsh.yaml
+++ b/target/_test/executable/test_workflows/ingestion/spaceranger_mapping_test/.config.vsh.yaml
@@ -48,7 +48,7 @@ repositories:
 - type: "vsh"
  name: "openpipeline"
  repo: "openpipeline"
-  tag: "v4.0.2"
+  tag: "v4.0.3"
 links:
  repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
  docker_registry: "ghcr.io"
@@ -157,7 +157,7 @@ build_info:
  output: "target/_test/executable/test_workflows/ingestion/spaceranger_mapping_test"
  executable: "target/_test/executable/test_workflows/ingestion/spaceranger_mapping_test/spaceranger_mapping_test"
  viash_version: "0.9.4"
-  git_commit: "ad19aba349736848d99a9646870d8e7868eb6515"
+  git_commit: "87e62605aafd706d539bf2978ef47ede6fe41926"
  git_remote: "https://github.com/openpipelines-bio/openpipeline_spatial"
 package_config:
  name: "openpipeline_spatial"
@@ -171,7 +171,7 @@ package_config:
  - type: "vsh"
    name: "openpipeline"
    repo: "openpipeline"
-    tag: "v4.0.2"
+    tag: "v4.0.3"
  viash_version: "0.9.4"
  source: "src"
  target: "target"
--- a/target/_test/executable/test_workflows/ingestion/spaceranger_mapping_test/spaceranger_mapping_test
+++ b/target/_test/executable/test_workflows/ingestion/spaceranger_mapping_test/spaceranger_mapping_test
@@ -459,9 +459,9 @@ RUN pip install --upgrade pip && \

 LABEL org.opencontainers.image.authors="Dorien Roosen"
 LABEL org.opencontainers.image.description="Companion container for running component test_workflows/ingestion spaceranger_mapping_test"
-LABEL org.opencontainers.image.created="2026-02-17T11:00:59Z"
+LABEL org.opencontainers.image.created="2026-03-25T10:11:23Z"
 LABEL org.opencontainers.image.source="https://github.com/openpipelines-bio/openpipeline_spatial"
-LABEL org.opencontainers.image.revision="ad19aba349736848d99a9646870d8e7868eb6515"
+LABEL org.opencontainers.image.revision="87e62605aafd706d539bf2978ef47ede6fe41926"
 LABEL org.opencontainers.image.version="niche-compass"

 VIASHDOCKER
--- a/target/_test/executable/test_workflows/niche/nichecompass_leiden_test/.config.vsh.yaml
+++ b/target/_test/executable/test_workflows/niche/nichecompass_leiden_test/.config.vsh.yaml
@@ -47,7 +47,7 @@ repositories:
 - type: "vsh"
  name: "openpipeline"
  repo: "openpipeline"
-  tag: "v4.0.2"
+  tag: "v4.0.3"
 links:
  repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
  docker_registry: "ghcr.io"
@@ -156,7 +156,7 @@ build_info:
  output: "target/_test/executable/test_workflows/niche/nichecompass_leiden_test"
  executable: "target/_test/executable/test_workflows/niche/nichecompass_leiden_test/nichecompass_leiden_test"
  viash_version: "0.9.4"
-  git_commit: "ad19aba349736848d99a9646870d8e7868eb6515"
+  git_commit: "87e62605aafd706d539bf2978ef47ede6fe41926"
  git_remote: "https://github.com/openpipelines-bio/openpipeline_spatial"
 package_config:
  name: "openpipeline_spatial"
@@ -170,7 +170,7 @@ package_config:
  - type: "vsh"
    name: "openpipeline"
    repo: "openpipeline"
-    tag: "v4.0.2"
+    tag: "v4.0.3"
  viash_version: "0.9.4"
  source: "src"
  target: "target"
--- a/target/_test/executable/test_workflows/niche/nichecompass_leiden_test/nichecompass_leiden_test
+++ b/target/_test/executable/test_workflows/niche/nichecompass_leiden_test/nichecompass_leiden_test
@@ -459,9 +459,9 @@ RUN pip install --upgrade pip && \

 LABEL org.opencontainers.image.authors="Dorien Roosen"
 LABEL org.opencontainers.image.description="Companion container for running component test_workflows/niche nichecompass_leiden_test"
-LABEL org.opencontainers.image.created="2026-02-17T11:00:59Z"
+LABEL org.opencontainers.image.created="2026-03-25T10:11:23Z"
 LABEL org.opencontainers.image.source="https://github.com/openpipelines-bio/openpipeline_spatial"
-LABEL org.opencontainers.image.revision="ad19aba349736848d99a9646870d8e7868eb6515"
+LABEL org.opencontainers.image.revision="87e62605aafd706d539bf2978ef47ede6fe41926"
 LABEL org.opencontainers.image.version="niche-compass"

 VIASHDOCKER
--- a/target/_test/nextflow/test_workflows/ingestion/spaceranger_mapping_test/.config.vsh.yaml
+++ b/target/_test/nextflow/test_workflows/ingestion/spaceranger_mapping_test/.config.vsh.yaml
@@ -48,7 +48,7 @@ repositories:
 - type: "vsh"
  name: "openpipeline"
  repo: "openpipeline"
-  tag: "v4.0.2"
+  tag: "v4.0.3"
 links:
  repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
  docker_registry: "ghcr.io"
@@ -157,7 +157,7 @@ build_info:
  output: "target/_test/nextflow/test_workflows/ingestion/spaceranger_mapping_test"
  executable: "target/_test/nextflow/test_workflows/ingestion/spaceranger_mapping_test/main.nf"
  viash_version: "0.9.4"
-  git_commit: "ad19aba349736848d99a9646870d8e7868eb6515"
+  git_commit: "87e62605aafd706d539bf2978ef47ede6fe41926"
  git_remote: "https://github.com/openpipelines-bio/openpipeline_spatial"
 package_config:
  name: "openpipeline_spatial"
@@ -171,7 +171,7 @@ package_config:
  - type: "vsh"
    name: "openpipeline"
    repo: "openpipeline"
-    tag: "v4.0.2"
+    tag: "v4.0.3"
  viash_version: "0.9.4"
  source: "src"
  target: "target"
--- a/target/_test/nextflow/test_workflows/ingestion/spaceranger_mapping_test/main.nf
+++ b/target/_test/nextflow/test_workflows/ingestion/spaceranger_mapping_test/main.nf
@@ -3104,7 +3104,7 @@ meta = [
      "type" : "vsh",
      "name" : "openpipeline",
      "repo" : "openpipeline",
-      "tag" : "v4.0.2"
+      "tag" : "v4.0.3"
    }
  ],
  "links" : {
@@ -3235,7 +3235,7 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/_test/nextflow/test_workflows/ingestion/spaceranger_mapping_test",
    "viash_version" : "0.9.4",
-    "git_commit" : "ad19aba349736848d99a9646870d8e7868eb6515",
+    "git_commit" : "87e62605aafd706d539bf2978ef47ede6fe41926",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline_spatial"
  },
  "package_config" : {
@@ -3255,7 +3255,7 @@ meta = [
        "type" : "vsh",
        "name" : "openpipeline",
        "repo" : "openpipeline",
-        "tag" : "v4.0.2"
+        "tag" : "v4.0.3"
      }
    ],
    "viash_version" : "0.9.4",
--- a/target/_test/nextflow/test_workflows/niche/nichecompass_leiden_test/.config.vsh.yaml
+++ b/target/_test/nextflow/test_workflows/niche/nichecompass_leiden_test/.config.vsh.yaml
@@ -47,7 +47,7 @@ repositories:
 - type: "vsh"
  name: "openpipeline"
  repo: "openpipeline"
-  tag: "v4.0.2"
+  tag: "v4.0.3"
 links:
  repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
  docker_registry: "ghcr.io"
@@ -156,7 +156,7 @@ build_info:
  output: "target/_test/nextflow/test_workflows/niche/nichecompass_leiden_test"
  executable: "target/_test/nextflow/test_workflows/niche/nichecompass_leiden_test/main.nf"
  viash_version: "0.9.4"
-  git_commit: "ad19aba349736848d99a9646870d8e7868eb6515"
+  git_commit: "87e62605aafd706d539bf2978ef47ede6fe41926"
  git_remote: "https://github.com/openpipelines-bio/openpipeline_spatial"
 package_config:
  name: "openpipeline_spatial"
@@ -170,7 +170,7 @@ package_config:
  - type: "vsh"
    name: "openpipeline"
    repo: "openpipeline"
-    tag: "v4.0.2"
+    tag: "v4.0.3"
  viash_version: "0.9.4"
  source: "src"
  target: "target"
--- a/target/_test/nextflow/test_workflows/niche/nichecompass_leiden_test/main.nf
+++ b/target/_test/nextflow/test_workflows/niche/nichecompass_leiden_test/main.nf
@@ -3104,7 +3104,7 @@ meta = [
      "type" : "vsh",
      "name" : "openpipeline",
      "repo" : "openpipeline",
-      "tag" : "v4.0.2"
+      "tag" : "v4.0.3"
    }
  ],
  "links" : {
@@ -3235,7 +3235,7 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/_test/nextflow/test_workflows/niche/nichecompass_leiden_test",
    "viash_version" : "0.9.4",
-    "git_commit" : "ad19aba349736848d99a9646870d8e7868eb6515",
+    "git_commit" : "87e62605aafd706d539bf2978ef47ede6fe41926",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline_spatial"
  },
  "package_config" : {
@@ -3255,7 +3255,7 @@ meta = [
        "type" : "vsh",
        "name" : "openpipeline",
        "repo" : "openpipeline",
-        "tag" : "v4.0.2"
+        "tag" : "v4.0.3"
      }
    ],
    "viash_version" : "0.9.4",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "split_modalities"
 namespace: "workflows/multiomics"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries Schaumont"
  roles:
@@ -183,13 +183,13 @@ build_info:
  output: "target/_private/nextflow/workflows/multiomics/split_modalities"
  executable: "target/_private/nextflow/workflows/multiomics/split_modalities/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
  dependencies:
  - "target/nextflow/dataflow/split_modalities"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -219,7 +219,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/main.nf
@@ -1,4 +1,4 @@
-// split_modalities v4.0.2
+// split_modalities v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "split_modalities",
  "namespace" : "workflows/multiomics",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries Schaumont",
@@ -3274,12 +3274,12 @@ meta = [
    "engine" : "native",
    "output" : "/workdir/root/repo/target/_private/nextflow/workflows/multiomics/split_modalities",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3304,7 +3304,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'workflows/multiomics/split_modalities'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'v4.0.2'
+  version = 'v4.0.3'
  description = 'A pipeline to split a multimodal mudata files into several unimodal mudata files.'
  author = 'Dries Schaumont'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/errorstrat_ignore.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/errorstrat_ignore.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/integration_tests.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/integration_tests.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/labels_ci.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/multiomics/split_modalities/utils/labels_ci.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "log_normalize"
 namespace: "workflows/rna"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries Schaumont"
  roles:
@@ -198,7 +198,7 @@ build_info:
  output: "target/_private/nextflow/workflows/rna/log_normalize"
  executable: "target/_private/nextflow/workflows/rna/log_normalize/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
  dependencies:
  - "target/nextflow/transform/normalize_total"
@@ -206,7 +206,7 @@ build_info:
  - "target/nextflow/transform/delete_layer"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -236,7 +236,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/main.nf
@@ -1,4 +1,4 @@
-// log_normalize v4.0.2
+// log_normalize v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "log_normalize",
  "namespace" : "workflows/rna",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries Schaumont",
@@ -3294,12 +3294,12 @@ meta = [
    "engine" : "native",
    "output" : "/workdir/root/repo/target/_private/nextflow/workflows/rna/log_normalize",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3324,7 +3324,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'workflows/rna/log_normalize'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'v4.0.2'
+  version = 'v4.0.3'
  description = 'Performs normalization and subsequent log-transformation of raw count data.'
  author = 'Dries Schaumont'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/errorstrat_ignore.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/errorstrat_ignore.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/integration_tests.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/integration_tests.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/labels_ci.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/_private/nextflow/workflows/rna/log_normalize/utils/labels_ci.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "leiden"
 namespace: "cluster"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries De Maeyer"
  roles:
@@ -215,7 +215,7 @@ engines:
  id: "docker"
  image: "python:3.13-slim"
  target_registry: "images.viash-hub.com"
-  target_tag: "v4.0.2"
+  target_tag: "v4.0.3"
  namespace_separator: "/"
  setup:
  - type: "apt"
@@ -258,11 +258,11 @@ build_info:
  output: "target/nextflow/cluster/leiden"
  executable: "target/nextflow/cluster/leiden/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -292,7 +292,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/compress_h5mu.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/compress_h5mu.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/main.nf
@@ -1,4 +1,4 @@
-// leiden v4.0.2
+// leiden v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "leiden",
  "namespace" : "cluster",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries De Maeyer",
@@ -3294,7 +3294,7 @@ meta = [
      "id" : "docker",
      "image" : "python:3.13-slim",
      "target_registry" : "images.viash-hub.com",
-      "target_tag" : "v4.0.2",
+      "target_tag" : "v4.0.3",
      "namespace_separator" : "/",
      "setup" : [
        {
@@ -3352,12 +3352,12 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/nextflow/cluster/leiden",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3382,7 +3382,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
@@ -4147,7 +4147,7 @@ meta["defaults"] = [
  "container" : {
    "registry" : "images.viash-hub.com",
    "image" : "vsh/openpipeline/cluster/leiden",
-    "tag" : "v4.0.2"
+    "tag" : "v4.0.3"
  },
  "label" : [
    "highcpu",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'cluster/leiden'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'v4.0.2'
+  version = 'v4.0.3'
  description = 'Cluster cells using the [Leiden algorithm] [Traag18] implemented in the [Scanpy framework] [Wolf18]. \nLeiden is an improved version of the [Louvain algorithm] [Blondel08]. \nIt has been proposed for single-cell analysis by [Levine15] [Levine15]. \nThis requires having ran `neighbors/find_neighbors` or `neighbors/bbknn` first.\n\n[Blondel08]: Blondel et al. (2008), Fast unfolding of communities in large networks, J. Stat. Mech.  \n[Levine15]: Levine et al. (2015), Data-Driven Phenotypic Dissection of AML Reveals Progenitor-like Cells that Correlate with Prognosis, Cell.  \n[Traag18]: Traag et al. (2018), From Louvain to Leiden: guaranteeing well-connected communities arXiv.  \n[Wolf18]: Wolf et al. (2018), Scanpy: large-scale single-cell gene expression data analysis, Genome Biology.  \n'
  author = 'Dries De Maeyer'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/nextflow_schema.json
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/nextflow_schema.json
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/setup_logger.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/cluster/leiden/setup_logger.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "concatenate_h5mu"
 namespace: "dataflow"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries Schaumont"
  roles:
@@ -253,7 +253,7 @@ engines:
  id: "docker"
  image: "python:3.13-slim"
  target_registry: "images.viash-hub.com"
-  target_tag: "v4.0.2"
+  target_tag: "v4.0.3"
  namespace_separator: "/"
  setup:
  - type: "apt"
@@ -300,11 +300,11 @@ build_info:
  output: "target/nextflow/dataflow/concatenate_h5mu"
  executable: "target/nextflow/dataflow/concatenate_h5mu/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -334,7 +334,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/compress_h5mu.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/compress_h5mu.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/main.nf
@@ -1,4 +1,4 @@
-// concatenate_h5mu v4.0.2
+// concatenate_h5mu v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "concatenate_h5mu",
  "namespace" : "dataflow",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries Schaumont",
@@ -3328,7 +3328,7 @@ meta = [
      "id" : "docker",
      "image" : "python:3.13-slim",
      "target_registry" : "images.viash-hub.com",
-      "target_tag" : "v4.0.2",
+      "target_tag" : "v4.0.3",
      "namespace_separator" : "/",
      "setup" : [
        {
@@ -3393,12 +3393,12 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/nextflow/dataflow/concatenate_h5mu",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3423,7 +3423,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
@@ -4266,7 +4266,7 @@ meta["defaults"] = [
  "container" : {
    "registry" : "images.viash-hub.com",
    "image" : "vsh/openpipeline/dataflow/concatenate_h5mu",
-    "tag" : "v4.0.2"
+    "tag" : "v4.0.3"
  },
  "label" : [
    "midcpu",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'dataflow/concatenate_h5mu'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'v4.0.2'
+  version = 'v4.0.3'
  description = 'Concatenate observations from samples in several (uni- and/or multi-modal) MuData files into a single file.\n'
  author = 'Dries Schaumont'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/nextflow_schema.json
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/nextflow_schema.json
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/setup_logger.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/concatenate_h5mu/setup_logger.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "merge"
 namespace: "dataflow"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries Schaumont"
  roles:
@@ -163,7 +163,7 @@ engines:
  id: "docker"
  image: "python:3.12-slim"
  target_registry: "images.viash-hub.com"
-  target_tag: "v4.0.2"
+  target_tag: "v4.0.3"
  namespace_separator: "/"
  setup:
  - type: "apt"
@@ -204,11 +204,11 @@ build_info:
  output: "target/nextflow/dataflow/merge"
  executable: "target/nextflow/dataflow/merge/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -238,7 +238,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/main.nf
@@ -1,4 +1,4 @@
-// merge v4.0.2
+// merge v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "merge",
  "namespace" : "dataflow",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries Schaumont",
@@ -3246,7 +3246,7 @@ meta = [
      "id" : "docker",
      "image" : "python:3.12-slim",
      "target_registry" : "images.viash-hub.com",
-      "target_tag" : "v4.0.2",
+      "target_tag" : "v4.0.3",
      "namespace_separator" : "/",
      "setup" : [
        {
@@ -3302,12 +3302,12 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/nextflow/dataflow/merge",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3332,7 +3332,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
@@ -3847,7 +3847,7 @@ meta["defaults"] = [
  "container" : {
    "registry" : "images.viash-hub.com",
    "image" : "vsh/openpipeline/dataflow/merge",
-    "tag" : "v4.0.2"
+    "tag" : "v4.0.3"
  },
  "label" : [
    "singlecpu",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'dataflow/merge'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'v4.0.2'
+  version = 'v4.0.3'
  description = 'Combine one or more single-modality .h5mu files together into one .h5mu file.\n'
  author = 'Dries Schaumont'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/nextflow_schema.json
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/nextflow_schema.json
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/setup_logger.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/merge/setup_logger.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "split_h5mu"
 namespace: "dataflow"
-version: "niche-compass"
+version: "v4.0.3"
 authors:
 - name: "Dorien Roosen"
  roles:
@@ -123,13 +123,9 @@ status: "enabled"
 scope:
  image: "public"
  target: "public"
-repositories:
- type: "vsh"
-  name: "openpipeline"
-  repo: "openpipeline"
-  tag: "v4.0.2"
+license: "MIT"
 links:
-  repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
+  repository: "https://github.com/openpipelines-bio/openpipeline"
  docker_registry: "ghcr.io"
 runners:
 - type: "executable"
@@ -207,7 +203,7 @@ engines:
  id: "docker"
  image: "python:3.12-slim"
  target_registry: "images.viash-hub.com"
-  target_tag: "niche-compass"
+  target_tag: "v4.0.3"
  namespace_separator: "/"
  setup:
  - type: "apt"
@@ -233,7 +229,7 @@ engines:
  - type: "python"
    user: false
    packages:
-    - "viashpy==0.9.0"
+    - "viashpy==0.8.0"
    github:
    - "openpipelines-bio/core#subdirectory=packages/python/openpipeline_testutils"
    upgrade: true
@@ -248,21 +244,31 @@ build_info:
  output: "target/nextflow/dataflow/split_h5mu"
  executable: "target/nextflow/dataflow/split_h5mu/main.nf"
  viash_version: "0.9.4"
-  git_commit: "ad19aba349736848d99a9646870d8e7868eb6515"
-  git_remote: "https://github.com/openpipelines-bio/openpipeline_spatial"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
+  git_remote: "https://github.com/openpipelines-bio/openpipeline"
 package_config:
-  name: "openpipeline_spatial"
-  version: "niche-compass"
+  name: "openpipeline"
+  version: "v4.0.3"
+  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
+  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
+    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
+    \nIn terms of workflows, the following has been made available, but keep in mind\
+    \ that\nindividual tools and functionality can be executed as standalone components\
+    \ as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n\
+    \  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample\
+    \ processing: cell filtering and doublet detection.\n  * Multisample processing:\
+    \ Count transformation, normalization, QC metric calulations.\n  * Integration:\
+    \ Clustering, integration and batch correction using single and multimodal methods.\n\
+    \  * Downstream analysis workflows\n"
  info:
    test_resources:
    - type: "s3"
-      path: "s3://openpipelines-bio/openpipeline_spatial/resources_test"
+      path: "s3://openpipelines-data"
      dest: "resources_test"
-  repositories:
-  - type: "vsh"
-    name: "openpipeline"
-    repo: "openpipeline"
-    tag: "v4.0.2"
+    nextflow_labels_ci:
+    - path: "src/workflows/utils/labels_ci.config"
+      description: "Adds the correct memory and CPU labels when running on the Viash\
+        \ Hub CI."
  viash_version: "0.9.4"
  source: "src"
  target: "target"
@@ -272,8 +278,15 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'niche-compass'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
+  keywords:
+  - "single-cell"
+  - "multimodal"
+  license: "MIT"
  organization: "vsh"
  links:
-    repository: "https://github.com/openpipelines-bio/openpipeline_spatial"
+    repository: "https://github.com/openpipelines-bio/openpipeline"
    docker_registry: "ghcr.io"
+    homepage: "https://openpipelines.bio"
+    documentation: "https://openpipelines.bio/fundamentals"
+    issue_tracker: "https://github.com/openpipelines-bio/openpipeline/issues"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/main.nf
@@ -1,4 +1,4 @@
-// split_h5mu niche-compass
+// split_h5mu v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "split_h5mu",
  "namespace" : "dataflow",
-  "version" : "niche-compass",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dorien Roosen",
@@ -3192,16 +3192,9 @@ meta = [
    "image" : "public",
    "target" : "public"
  },
-  "repositories" : [
-    {
-      "type" : "vsh",
-      "name" : "openpipeline",
-      "repo" : "openpipeline",
-      "tag" : "v4.0.2"
-    }
-  ],
+  "license" : "MIT",
  "links" : {
-    "repository" : "https://github.com/openpipelines-bio/openpipeline_spatial",
+    "repository" : "https://github.com/openpipelines-bio/openpipeline",
    "docker_registry" : "ghcr.io"
  },
  "runners" : [
@@ -3292,7 +3285,7 @@ meta = [
      "id" : "docker",
      "image" : "python:3.12-slim",
      "target_registry" : "images.viash-hub.com",
-      "target_tag" : "niche-compass",
+      "target_tag" : "v4.0.3",
      "namespace_separator" : "/",
      "setup" : [
        {
@@ -3328,7 +3321,7 @@ meta = [
          "type" : "python",
          "user" : false,
          "packages" : [
-            "viashpy==0.9.0"
+            "viashpy==0.8.0"
          ],
          "github" : [
            "openpipelines-bio/core#subdirectory=packages/python/openpipeline_testutils"
@@ -3348,29 +3341,29 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/nextflow/dataflow/split_h5mu",
    "viash_version" : "0.9.4",
-    "git_commit" : "ad19aba349736848d99a9646870d8e7868eb6515",
-    "git_remote" : "https://github.com/openpipelines-bio/openpipeline_spatial"
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
+    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
-    "name" : "openpipeline_spatial",
-    "version" : "niche-compass",
+    "name" : "openpipeline",
+    "version" : "v4.0.3",
+    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
+    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
      "test_resources" : [
        {
          "type" : "s3",
-          "path" : "s3://openpipelines-bio/openpipeline_spatial/resources_test",
+          "path" : "s3://openpipelines-data",
          "dest" : "resources_test"
        }
+      ],
+      "nextflow_labels_ci" : [
+        {
+          "path" : "src/workflows/utils/labels_ci.config",
+          "description" : "Adds the correct memory and CPU labels when running on the Viash Hub CI."
+        }
      ]
    },
-    "repositories" : [
-      {
-        "type" : "vsh",
-        "name" : "openpipeline",
-        "repo" : "openpipeline",
-        "tag" : "v4.0.2"
-      }
-    ],
    "viash_version" : "0.9.4",
    "source" : "/workdir/root/repo/src",
    "target" : "/workdir/root/repo/target",
@@ -3378,12 +3371,20 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'niche-compass'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
+    "keywords" : [
+      "single-cell",
+      "multimodal"
+    ],
+    "license" : "MIT",
    "organization" : "vsh",
    "links" : {
-      "repository" : "https://github.com/openpipelines-bio/openpipeline_spatial",
-      "docker_registry" : "ghcr.io"
+      "repository" : "https://github.com/openpipelines-bio/openpipeline",
+      "docker_registry" : "ghcr.io",
+      "homepage" : "https://openpipelines.bio",
+      "documentation" : "https://openpipelines.bio/fundamentals",
+      "issue_tracker" : "https://github.com/openpipelines-bio/openpipeline/issues"
    }
  }
 }'''))
@@ -3908,8 +3909,8 @@ meta["defaults"] = [
  directives: readJsonBlob('''{
  "container" : {
    "registry" : "images.viash-hub.com",
-    "image" : "vsh/openpipeline_spatial/dataflow/split_h5mu",
-    "tag" : "niche-compass"
+    "image" : "vsh/openpipeline/dataflow/split_h5mu",
+    "tag" : "v4.0.3"
  },
  "label" : [
    "lowcpu",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'dataflow/split_h5mu'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'niche-compass'
+  version = 'v4.0.3'
  description = 'Split the samples of a single modality from a .h5mu (multimodal) sample into seperate .h5mu files based on the values of an .obs column of this modality. \n'
  author = 'Dorien Roosen'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dataflow/split_modalities/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dataflow/split_modalities/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/nextflow_schema.json
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_h5mu/nextflow_schema.json
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dataflow/split_modalities/setup_logger.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dataflow/split_modalities/setup_logger.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "split_modalities"
 namespace: "dataflow"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries Schaumont"
  roles:
@@ -190,7 +190,7 @@ engines:
  id: "docker"
  image: "python:3.12-slim"
  target_registry: "images.viash-hub.com"
-  target_tag: "v4.0.2"
+  target_tag: "v4.0.3"
  namespace_separator: "/"
  setup:
  - type: "apt"
@@ -231,11 +231,11 @@ build_info:
  output: "target/nextflow/dataflow/split_modalities"
  executable: "target/nextflow/dataflow/split_modalities/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -265,7 +265,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/main.nf
@@ -1,4 +1,4 @@
-// split_modalities v4.0.2
+// split_modalities v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3036,7 +3036,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "split_modalities",
  "namespace" : "dataflow",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries Schaumont",
@@ -3280,7 +3280,7 @@ meta = [
      "id" : "docker",
      "image" : "python:3.12-slim",
      "target_registry" : "images.viash-hub.com",
-      "target_tag" : "v4.0.2",
+      "target_tag" : "v4.0.3",
      "namespace_separator" : "/",
      "setup" : [
        {
@@ -3336,12 +3336,12 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/nextflow/dataflow/split_modalities",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3366,7 +3366,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
@@ -3865,7 +3865,7 @@ meta["defaults"] = [
  "container" : {
    "registry" : "images.viash-hub.com",
    "image" : "vsh/openpipeline/dataflow/split_modalities",
-    "tag" : "v4.0.2"
+    "tag" : "v4.0.3"
  },
  "label" : [
    "singlecpu",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'dataflow/split_modalities'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'v4.0.2'
+  version = 'v4.0.3'
  description = 'Split the modalities from a single .h5mu multimodal sample into seperate .h5mu files. \n'
  author = 'Dries Schaumont, Robrecht Cannoodt'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/nextflow_schema.json
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/nextflow_schema.json
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/setup_logger.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dataflow/split_modalities/setup_logger.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "pca"
 namespace: "dimred"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries De Maeyer"
  roles:
@@ -240,7 +240,7 @@ engines:
  id: "docker"
  image: "python:3.12-slim"
  target_registry: "images.viash-hub.com"
-  target_tag: "v4.0.2"
+  target_tag: "v4.0.3"
  namespace_separator: "/"
  setup:
  - type: "apt"
@@ -276,11 +276,11 @@ build_info:
  output: "target/nextflow/dimred/pca"
  executable: "target/nextflow/dimred/pca/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -310,7 +310,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/compress_h5mu.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/compress_h5mu.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/main.nf
@@ -1,4 +1,4 @@
-// pca v4.0.2
+// pca v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "pca",
  "namespace" : "dimred",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries De Maeyer",
@@ -3333,7 +3333,7 @@ meta = [
      "id" : "docker",
      "image" : "python:3.12-slim",
      "target_registry" : "images.viash-hub.com",
-      "target_tag" : "v4.0.2",
+      "target_tag" : "v4.0.3",
      "namespace_separator" : "/",
      "setup" : [
        {
@@ -3380,12 +3380,12 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/nextflow/dimred/pca",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3410,7 +3410,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
@@ -3933,7 +3933,7 @@ meta["defaults"] = [
  "container" : {
    "registry" : "images.viash-hub.com",
    "image" : "vsh/openpipeline/dimred/pca",
-    "tag" : "v4.0.2"
+    "tag" : "v4.0.3"
  },
  "label" : [
    "highcpu",
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/nextflow.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/nextflow.config
@@ -2,7 +2,7 @@ manifest {
  name = 'dimred/pca'
  mainScript = 'main.nf'
  nextflowVersion = '!>=20.12.1-edge'
-  version = 'v4.0.2'
+  version = 'v4.0.3'
  description = 'Computes PCA coordinates, loadings and variance decomposition. Uses the implementation of scikit-learn [Pedregosa11].\n'
  author = 'Dries De Maeyer'
 }
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dimred/umap/nextflow_labels.config
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dimred/umap/nextflow_labels.config
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/nextflow_schema.json
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/pca/nextflow_schema.json
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dimred/umap/setup_logger.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.2/nextflow/dimred/umap/setup_logger.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/umap/.config.vsh.yaml
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/umap/.config.vsh.yaml
@@ -1,6 +1,6 @@
 name: "umap"
 namespace: "dimred"
-version: "v4.0.2"
+version: "v4.0.3"
 authors:
 - name: "Dries De Maeyer"
  roles:
@@ -294,7 +294,7 @@ engines:
  id: "docker"
  image: "python:3.12-slim"
  target_registry: "images.viash-hub.com"
-  target_tag: "v4.0.2"
+  target_tag: "v4.0.3"
  namespace_separator: "/"
  setup:
  - type: "apt"
@@ -330,11 +330,11 @@ build_info:
  output: "target/nextflow/dimred/umap"
  executable: "target/nextflow/dimred/umap/main.nf"
  viash_version: "0.9.4"
-  git_commit: "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62"
+  git_commit: "7bfad4ea12f87eca59213be3ab08deff67cc4206"
  git_remote: "https://github.com/openpipelines-bio/openpipeline"
 package_config:
  name: "openpipeline"
-  version: "v4.0.2"
+  version: "v4.0.3"
  summary: "Best-practice workflows for single-cell multi-omics analyses.\n"
  description: "OpenPipelines are extensible single cell analysis pipelines for reproducible\
    \ and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\
@@ -364,7 +364,7 @@ package_config:
    )'"
  - ".engines += { type: \"native\" }"
  - ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'"
-  - ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+  - ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
  keywords:
  - "single-cell"
  - "multimodal"
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/umap/compress_h5mu.py
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/umap/compress_h5mu.py
--- a/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/umap/main.nf
+++ b/target/dependencies/vsh/vsh/openpipeline/v4.0.3/nextflow/dimred/umap/main.nf
@@ -1,4 +1,4 @@
-// umap v4.0.2
+// umap v4.0.3
 // 
 // This wrapper script is auto-generated by viash 0.9.4 and is thus a derivative
 // work thereof. This software comes with ABSOLUTELY NO WARRANTY from Data
@@ -3035,7 +3035,7 @@ meta = [
  "config": processConfig(readJsonBlob('''{
  "name" : "umap",
  "namespace" : "dimred",
-  "version" : "v4.0.2",
+  "version" : "v4.0.3",
  "authors" : [
    {
      "name" : "Dries De Maeyer",
@@ -3382,7 +3382,7 @@ meta = [
      "id" : "docker",
      "image" : "python:3.12-slim",
      "target_registry" : "images.viash-hub.com",
-      "target_tag" : "v4.0.2",
+      "target_tag" : "v4.0.3",
      "namespace_separator" : "/",
      "setup" : [
        {
@@ -3429,12 +3429,12 @@ meta = [
    "engine" : "docker|native",
    "output" : "/workdir/root/repo/target/nextflow/dimred/umap",
    "viash_version" : "0.9.4",
-    "git_commit" : "cd7beddfa85b60b543399a9e2554b8e7b8eaaf62",
+    "git_commit" : "7bfad4ea12f87eca59213be3ab08deff67cc4206",
    "git_remote" : "https://github.com/openpipelines-bio/openpipeline"
  },
  "package_config" : {
    "name" : "openpipeline",
-    "version" : "v4.0.2",
+    "version" : "v4.0.3",
    "summary" : "Best-practice workflows for single-cell multi-omics analyses.\n",
    "description" : "OpenPipelines are extensible single cell analysis pipelines for reproducible and large-scale single cell processing using [Viash](https://viash.io) and [Nextflow](https://www.nextflow.io/).\n\nIn terms of workflows, the following has been made available, but keep in mind that\nindividual tools and functionality can be executed as standalone components as well.\n\n  * Demultiplexing: conversion of raw sequencing data to FASTQ objects.\n  * Ingestion: Read mapping and generating a count matrix.\n  * Single sample processing: cell filtering and doublet detection.\n  * Multisample processing: Count transformation, normalization, QC metric calulations.\n  * Integration: Clustering, integration and batch correction using single and multimodal methods.\n  * Downstream analysis workflows\n",
    "info" : {
@@ -3459,7 +3459,7 @@ meta = [
      ".resources += {path: '/src/workflows/utils/labels.config', dest: 'nextflow_labels.config'}\n.runners[.type == 'nextflow'].config.script := 'includeConfig(\\"nextflow_labels.config\\")'",
      ".engines += { type: \\"native\\" }",
      ".engines[.type == 'docker'].target_registry := 'images.viash-hub.com'",
-      ".engines[.type == 'docker'].target_tag := 'v4.0.2'"
+      ".engines[.type == 'docker'].target_tag := 'v4.0.3'"
    ],
    "keywords" : [
      "single-cell",
@@ -3971,7 +3971,7 @@ meta["defaults"] = [
  "container" : {
    "registry" : "images.viash-hub.com",
    "image" : "vsh/openpipeline/dimred/umap",
-    "tag" : "v4.0.2"
+    "tag" : "v4.0.3"
  },
  "label" : [
    "highcpu",
--- a/Show More
+++ b/Show More