Skip to content

Installation

GenAIRR ships as a single pre-built wheel that bundles both the Python API and the Rust simulation kernel. No compiler is required on any supported platform.

Requirements

  • Python 3.9 or newer (3.9, 3.10, 3.11, 3.12, and 3.13 are officially supported and CI-tested).
  • One of the following platforms with a pre-built wheel:
    • Linux x86_64 or aarch64
    • macOS Intel or Apple Silicon
    • Windows x64

If your platform isn't covered, the source build needs a stable Rust toolchain (rustup install stable); see CONTRIBUTING.md.

Install

pip install GenAIRR

That's the whole install. Nothing else needs to be on your system — no Rust toolchain, no compiler, no external services. A virtual environment (venv or conda) is recommended to avoid clashing with other repertoire-analysis packages.

Smoke test

After install, a one-liner confirms everything is in place:

import GenAIRR as ga

print(ga.__version__)

If this prints a version string, the Python API loaded and the Rust kernel was importable. To run a 5-record simulation as a deeper smoke:

result = ga.Experiment.on("human_igh").recombine().run_records(n=5, seed=0)
assert len(result) == 5
print(result[0]["v_call"], result[0]["junction_aa"])

Optional extras

A few features live behind opt-in extras to keep the base wheel small:

pip install GenAIRR[all]          # numpy, scipy, graphviz, tqdm, fastmcp
pip install GenAIRR[dataconfig]   # numpy + scipy (custom DataConfig analysis)
pip install GenAIRR[viz]          # graphviz
pip install GenAIRR[mcp]          # fastmcp (for the MCP server)

pandas is needed for result.to_dataframe(); it ships under [all]. The core simulator works without any extras — every export format except to_dataframe is implemented in pure Python on the standard library.

Documentation build (optional, for contributors)

To build this documentation site locally:

pip install -r requirements-docs.txt
make docs-serve

The site then renders at http://localhost:8000 with live reload. make docs-build runs a --strict build that fails on broken links — the same command CI uses.

Next step

→ First simulation — generate 1,000 productive heavy-chain sequences and inspect what comes back.