scikit-bio at Google Summer of Code 2026

[qiyunzhu]

We are excited to announce that scikit-bio is participating in Google Summer of Code (GSoC) 2026 as part of the NumFOCUS umbrella organization.

This year, we aim to modernize fundamental bioinformatics algorithms in scikit-bio to support GPU acceleration and interoperability across AI-ready array libraries.

Many core algorithms in scikit-bio, such as those for ordination, diversity analysis, and compositional data analysis, were originally implemented on top of NumPy and SciPy. Recent advances such as the Python array API standard, the xp namespace, and Numba, now make it possible to write backend-agnostic numerical code that can run on multiple array libraries, such as CuPy, PyTorch, JAX and Dask. Check out our project ideas:

• https://github.com/scikit-bio/scikit-bio/wiki/GSOC-2026-project-ideas

If you are interested in scientific Python, bioinformatics, or GPU computing, we encourage you to explore the scikit-bio repo and start engaging! Feel free to rely in this thread, open issues / pull requests, or email us (qiyunzhu@gmail.com).

We look forward to welcoming new open source contributors this summer!

[MAUK9086]

Hi @qiyunzhu and team,

I'm interested in the Array API modernization project. I’ve recently had 2-3 PRs merged in scikit-learn (some awaiting review) , so I'm familiar with the codebase standards and performance requirements of the scientific Python stack.

I’m comfortable working with the xp namespace and ensuring numerical consistency across backends. I saw that Alpha diversity and CoDa transformations (CLR/ALR) are listed as priorities. Do you have a specific function or module you'd like me to attempt a prototype refactor on as a qualification task?

[qiyunzhu]

Hello Mohammad @MAUK9086 thanks for your interest! I saw your PRs in scikit-learn. You are welcome to contribute to scikit-bio. As you might see, there are some recent community PRs attempting to address alpha diversity and CoDa transformations. You may instead navigate the issue tracker to identify unaddressed issues. Or, you may try to run some scikit-bio functions yourself to get a feeling, and identify things that can be improved.

[Naty89]

Hi @qiyunzhu ! I'm Atnatiwos Fantahun, a Computational Biology student at RPI applying for GSoC 2026, I am very interested in this project as well! I have some direct experiences with several of the algorithms being targets, specifiically the differential abundance testing DESeq2 and ordination methods in genomics research, so I have an understanding of what these function are supposed to do biologically. Following your suggestion, I am gonna explore the issue tracker and rum some of the alpha diversity and CoDA functions and look for improvement areas. Is there a preferred array backend to prioritize first first?