rapids-singlecell: GPU-Accelerated Single-Cell Analysis within scverse®

rapids-singlecell provides GPU-accelerated single-cell analysis with an AnnData-first API. It is largely compatible with Scanpy and includes selected functionality from Squidpy, decoupler, and pertpy. Computations use CuPy and NVIDIA CUDA-X Data Science (formerly RAPIDS) for performance on large datasets.

• GPU acceleration: Common single-cell workflows on AnnData run on the GPU.
• Ecosystem compatibility: Works with Scanpy APIs; includes pieces from Squidpy, decoupler, and pertpy.
• Simple installation: Available via Conda and PyPI.
• Version-matched agent skill: Released pip packages include an analysis skill for Codex, Claude, and other agents.

Agent skill

Released packages include the matching model-neutral analysis skill:

rapids-singlecell-install-skills --agent codex
rapids-singlecell-install-skills --agent claude
rapids-singlecell-install-skills --agent claude-science  # or: --agent agents
rapids-singlecell-check-kernel                    # before starting Jupyter

Use --dest /path/to/skills/rapids-singlecell for another agent. See the installation guide for installation and preflight details.

Documentation

For more information please have a look through the documentation

Citation

If you use this tool, please cite:

Please cite the relevant tools if used: decoupler for decoupler functions, squidpy for spatial analysis, and pertpy for perturbation analysis.

rapids-singlecell is part of the scverse® project (website, governance) and is fiscally sponsored by NumFOCUS. If you like scverse® and want to support our mission, please consider making a tax-deductible donation to help the project pay for developer time, professional services, travel, workshops, and a variety of other needs.