tiledbsoma

What it is and what it does

TileDB-SOMA provides a Python API for the Unified Single-cell Data Model (SOMA), an open standard for organizing and accessing single-cell genomic data. It uses TileDB, a columnar array storage engine, as its backend to enable efficient storage and retrieval of large single-cell datasets. The package implements the full SOMA specification, allowing researchers and bioinformaticians to work with standardized single-cell data formats across different tools and platforms.

The package integrates with the popular single-cell Python ecosystem—it depends on anndata, scanpy, pandas, numpy, and pyarrow—making it a bridge between TileDB's storage capabilities and existing bioinformatics workflows. It supports platform-specific configuration through a TypeScript-style interface, allowing fine-grained control over TileDB storage parameters like filters, cell order, and capacity settings.

Use it for:

• Store and query large single-cell RNA-seq datasets in a standardized, interoperable format using TileDB's efficient columnar storage.
• Integrate single-cell data workflows with tools like scanpy and anndata while maintaining SOMA compliance across different analysis pipelines.
• Configure TileDB-specific storage options (filters, dimensions, compression) for optimized performance on custom single-cell datasets.
• Access single-cell data through a standardized API that abstracts away storage backend details, enabling portability between implementations.
• Build reproducible bioinformatics pipelines that rely on a unified data model for single-cell genomics across multiple research groups.

Worth the install?

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

TileDB-SOMA is a Python implementation of the SOMA API specification for storing and retrieving single-cell genomic data using TileDB as the underlying storage engine.

Yes, if you work with single-cell genomic data and need standardized, efficient storage. The package is actively maintained, supports modern Python versions (3.9–3.13), has no known vulnerabilities, and integrates well with the established scanpy and anndata ecosystem. Medium install friction is manageable for most development environments. Not necessary if you are already satisfied with your current single-cell data storage and don't require SOMA compliance or TileDB's specific performance characteristics.

Install

tiledbsoma on PyPI

pip install tiledbsoma

uv add tiledbsoma

poetry add tiledbsoma

Installing tiledbsoma

Before you install

Medium install friction due to compiled binary wheels for multiple Python versions and architectures. Active maintenance with recent releases. Requires 10 runtime dependencies including numpy, pandas, pyarrow, and scanpy. Pre-built wheels available for macOS and Linux; source installation may be needed on older processors lacking AVX2 support.

License in practice

MIT license permits commercial and private use with minimal restrictions, making it suitable for both academic and production environments.

Quickstart

pip install tiledbsoma

import tiledbsoma as soma
# Access SOMA data via TileDB backend with platform_config for TileDB-specific settings

Requires Python 3.9 or later. On older CPU architectures without AVX2 support, source installation from the repository may be necessary instead of using pre-compiled wheels.

Verify before relying

• Whether the package requires a TileDB server or library installation separate from the Python package itself
• Performance characteristics and scalability limits for typical single-cell datasets
• Compatibility with specific versions of scanpy, anndata, or other bioinformatics tools in the dependency chain

Package facts

Evidence: tiledbsoma-2.3.0-cp310-cp310-macosx_13_0_arm64.whl; tiledbsoma-2.3.0-cp310-cp310-macosx_13_0_x86_64.whl; tiledbsoma-2.3.0-cp310-cp310-manylinux_2_28_aarch64.whl; tiledbsoma-2.3.0-cp310-cp310-manylinux_2_28_x86_64.whl; tiledbsoma-2.3.0-cp311-cp311-macosx_13_0_arm64.whl; tiledbsoma-2.3.0-cp311-cp311-macosx_13_0_x86_64.whl; tiledbsoma-2.3.0-cp311-cp311-manylinux_2_28_aarch64.whl; tiledbsoma-2.3.0-cp311-cp311-manylinux_2_28_x86_64.whl; tiledbsoma-2.3.0-cp312-cp312-macosx_13_0_arm64.whl; tiledbsoma-2.3.0-cp312-cp312-macosx_13_0_x86_64.whl; tiledbsoma-2.3.0-cp312-cp312-manylinux_2_28_aarch64.whl; tiledbsoma-2.3.0-cp312-cp312-manylinux_2_28_x86_64.whl; tiledbsoma-2.3.0-cp313-cp313-macosx_13_0_arm64.whl; tiledbsoma-2.3.0-cp313-cp313-macosx_13_0_x86_64.whl; tiledbsoma-2.3.0-cp313-cp313-manylinux_2_28_aarch64.whl; tiledbsoma-2.3.0-cp313-cp313-manylinux_2_28_x86_64.whl; tiledbsoma-2.3.0-cp39-cp39-macosx_13_0_arm64.whl; tiledbsoma-2.3.0-cp39-cp39-macosx_13_0_x86_64.whl; tiledbsoma-2.3.0-cp39-cp39-manylinux_2_28_aarch64.whl; tiledbsoma-2.3.0-cp39-cp39-manylinux_2_28_x86_64.whl

Tags

More Bio-Informatics packages