spatialdata

What it is and what it does

SpatialData is a framework and schema for storing and working with spatial omics data—datasets that combine gene expression, protein abundance, or other molecular measurements with spatial coordinates. It defines a universal format built on the OME-NGFF specification, allowing researchers to represent uni- and multi-modal spatial experiments in a standardized way. The core library handles data access, manipulation, and serialization; it is part of a growing ecosystem that includes companion packages for loading data from common technologies, static plotting, and interactive exploration.

The package depends on a large stack of scientific libraries—including dask for distributed computing, geopandas for spatial operations, xarray for multi-dimensional arrays, and zarr for chunked storage—making it suitable for workflows that need to process, analyze, or share spatial omics datasets across teams or languages. It is actively maintained by the scverse project and has been cited in peer-reviewed literature, though the documentation notes that the library is still under community review and may see API changes.

Use it for:

• Load spatial transcriptomics or proteomics data from commercial platforms (e.g., Visium, MERFISH) into a standardized format for downstream analysis.
• Store and share multi-modal spatial omics experiments (combining imaging, gene expression, and metadata) in a single interoperable file.
• Build analysis pipelines that work with spatial data in Python while ensuring compatibility with R and JavaScript implementations.
• Perform distributed spatial operations (clustering, segmentation, neighborhood analysis) on large datasets using dask integration.
• Visualize spatial omics data interactively via the napari plugin or generate publication-ready static plots.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

SpatialData is a data framework and serialization format for uni- and multi-modal spatial omics datasets, providing schema-based storage and access across Python, R, and JavaScript implementations.

Yes, if you work with spatial omics data and need a standardized, cross-language framework. The package is actively maintained, permissively licensed, and backed by a peer-reviewed publication and institutional support. Install friction is low. However, be aware that it requires Python 3.12, carries substantial downstream dependencies, and the documentation explicitly notes the library is under review—expect potential API changes as the community provides feedback. Not suitable if you need Windows-specific support guarantees or absolute API stability.

Install

spatialdata on PyPI

pip install spatialdata

uv add spatialdata

poetry add spatialdata

Installing spatialdata

Before you install

Low install friction with a pure-wheel distribution. Active maintenance with a recent release (43 days old) and ongoing development. The package carries 28 runtime dependencies including heavy scientific stacks (dask, geopandas, xarray, zarr), which will pull in substantial downstream requirements but are all standard in the spatial-omics ecosystem.

License in practice

BSD 3-Clause permissive license allows commercial and private use with minimal restrictions; you must retain copyright notice and disclaimer in distributions.

Quickstart

pip install spatialdata

import spatialdata
# Load or create spatial omics data via the framework
data = spatialdata.read_zarr('path/to/data.zarr')

Requires Python 3.12 or later. Installation via conda is currently not available; pip is the only supported installation method.

Verify before relying

• Specific performance characteristics or scalability limits for large spatial datasets.
• Maturity level and API stability guarantees given the note that the library is 'currently under review'.
• Windows support status beyond 'manually tested' — automated testing covers Linux and macOS only.

Package facts

Evidence: spatialdata-0.8.0-py3-none-any.whl

Tags

More Information Analysis packages