Installing scikit-bio — scikit-bio

Scikit-bio can be installed via PyPI, Conda, system package managers, from source, or from pre-compiled wheels. The PyPI and Conda methods are the most commonly used.
Scikit-bio runs on Linux, macOS, and Windows. It is currently available for the x86_64 and ARM64 architectures.
scikit-bio requires Python 3.11 or later installed in your system. See the Python version support section for more details.
The recommended way to install scikit-bio is via the Conda package manager. The latest release of scikit-bio is distributed via the conda-forge channel. You can install it via the following command:
Other channels such as anaconda and bioconda also host scikit-bio, which however may or may not be the up-to-date version.
Alternatively, the latest release of scikit-bio can be installed from PyPI:
Scikit-bio is available as third-party packages from software repositories for multiple Linux/BSD distributions. However, these packages may or may not be the latest version. The scikit-bio development team is not involved in the maintenance of these packages.
For example, users of Debian-based Linux distributions (such as Ubuntu and Linux Mint) may install scikit-bio using:
Users of Arch Linux or variants (such as Manjaro) may install scikit-bio from AUR:
Starting with version 0.7.0, scikit-bio now provides pre-compiled wheels for each release. To install from a wheel file, download the appropriate file for your platform and run:
Scikit-bio is undergoing expansion, with many new features being introduced. You are welcome to try these features by installing the current development version from our GitHub repo:
Alternatively, you may download the repository, extract, and execute:
However, be cautious that the new functionality may not be stable and could be changed in the next formal release. It is not recommended to deploy the development version in a production environment.
After installing scikit-bio, verify the installation by running the following in a Python shell or script:
This should print the installed version of scikit-bio without errors.
For a more robust verification, install pytest and run scikit-bio’s unit tests in the same environment:
Tests requiring optional packages such as Matplotlib are skipped when those packages are absent. To run the full test suite, install the test dependencies instead:
Multiple compute-intensive algorithms in scikit-bio are implemented in Cython and use OpenMP for parallel execution. If your system does not support OpenMP, or if you prefer to disable it, you can build from the scikit-bio source code with OpenMP disabled:
Or, if you have downloaded the repository:
For runtime thread controls, including OpenMP, Numba, and numerical libraries, see the parallelization guide.
Some scikit-bio functions offer an alternative Numba engine to replace the default Python/Cython engine for efficient computing. To utilize the Numba engine, install Numba with:
See the compute engine guide for details.
Some functions’ Numba engines support GPU computing if available. To enable GPU computing via Numba, you need to install an architecture-specific Numba extension that matches your GPU device:
The older NVIDIA extension, numba-cuda, is still supported but is deprecated and requires NumPy 2.4 or earlier.
see the GPU computing guide. for details.
scikit-bio-binaries is a separate package from scikit-bio, written in C++, which when installed in the same environment as scikit-bio will dramatically increase performance of select functions. Installation of scikit-bio-binaries is currently available with conda:
For applicability and interaction with compute engines, see the computation and performance guide.
Scikit-bio’s policy is to support Python versions until they reach official end-of-life, which typically provides 4-5 years of support per version. This approach accounts for the prevalence of legacy and under-maintained software packages in bioinformatics workflows, where users often cannot update their environments frequently. The goal is to strike an appropriate balance between encouraging modernization and maintaining compatibility for our user base, while still providing predictable deprecation timelines that researchers can plan around.
© Copyright 2014-2026, scikit-bio development team.
Created using Sphinx 9.2.0.
Built with the PyData Sphinx Theme 0.24.0.