Mantpy: extracellular-matrix analysis for spatial proteomics

Mantpy is a scverse-based framework for graph analysis of the extracellular matrix (ECM) in spatial proteomics. It represents cells and ECM patches as distinct, linked node types so that matrix structure can be analysed on its own or together with cellular context. Mantpy works with AnnData and interoperates with Scanpy, Squidpy, and other single-cell and spatial tools.

Installation

Mantpy requires Python 3.11 or newer. Install it from PyPI with:

pip install mantpy

Optional extras add heavier dependencies only when needed:

pip install "mantpy[gnn]"      # graph learning and explainability
pip install "mantpy[patch]"    # learned image-patch features
pip install "mantpy[spatial]"  # SpatialData integration
pip install "mantpy[segment]"  # Cellpose segmentation

For development from a source checkout, replace mantpy with . in those commands.

Quick start

import mantpy as mt

# Read multiplexed imaging and a cell table into AnnData.
adata = mt.io.read_imc("image.tiff", panel="panel.csv", cells="cells.csv")

# Normalise channels and segment the ECM into patch nodes.
mt.pp.normalize(adata)
mt.pp.extract_ecm_patches(
    adata,
    ecm_channel="Collagen",
    ecm_K="auto",
    features=["mean"],
)

# Build cell, ECM, and joint cell-ECM graph layers.
mt.gr.build_graph(adata, mode="cell")
mt.gr.build_graph(adata, mode="ecm")
mt.gr.build_graph(adata, mode="cell_ecm")

# Quantify spatial organisation.
mt.tl.cell_ecm_enrichment(adata, cell_type="Macrophage")
mt.tl.neighbourhood_clustering(adata, n_clusters=4)

# Visualise results.
mt.pl.cell_ecm_graph(adata)
mt.pl.neighbourhood_clusters(adata)

Public tutorial data

The complete inputs for each worked tutorial are available through one-line, checksummed loaders from the immutable Zenodo version record. Data are cached outside the package and reused offline:

intestine = mt.datasets.coliv_intestine()
lung = mt.datasets.balbc_pbs_lung()
liver = mt.datasets.schistosoma_ecm()
prostate = mt.datasets.prostate_he_visium()

AnnData and image containers

Mantpy stores the image-container payload in an H5AD-safe form. Convert that payload back to the object interface whenever you need direct image access:

image = mt.im.as_image_container(adata.uns["image_container"])
image.layers

The conversion accepts both a serialized mapping restored from H5AD and a live mt.im.ImageContainer created in memory.

Plot styling

Importing Mantpy does not change Matplotlib defaults. Apply the generic, export-friendly preset explicitly when desired:

mt.style.apply_publication_style()

Features

• Graph-based ECM modelling with cell, ECM, and joint cell-ECM layers
• Cell-ECM enrichment and interaction testing
• Spatial neighbourhood clustering
• Label-free spatial-domain discovery
• Optional graph embeddings, node classification, and denoising
• Native AnnData interoperability throughout the workflow

API overview

The final worked tutorials are maintained in the Mantpy reproducibility repository. See the documentation and API documentation for complete signatures and examples.

Release notes

See the changelog.

Contact

For questions and help requests, use the scverse discourse. To report a bug, use the issue tracker.