# What changed in scikit-bio

> 30 changes in scikit-bio 0.7 from its official release notes, upgrade guides and advisories: 15 new APIs and options. Each one cited. Sources checked 2026-10-30. HTML: https://unstale.sh/pypi/scikit-bio/

To see which of these apply to a repository, run `npx @unstale/cli init` in it.

## scikit-bio 0.7

### New APIs and options

- Adds an efficient Python implementation of ANCOM-BC2 for differential abundance analysis, including post-hoc tests. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds a multiple sequence alignment function using the classical progressive alignment method. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds optional Numba acceleration for PCoA. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds the inverse robust centered log-ratio transformation. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds transition probability matrices for five nucleotide substitution models. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds optional Numba acceleration for UniFrac. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Added gamma correction for among-site evolutionary rate heterogeneity for multiple nucleotide substitution models. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- Added Microbe-Metabolite Vectors for multiomics co-occurrence analysis and embedding; the implementation is significantly faster than the original program and does not require GPU. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- A global mechanism was implemented to support the Python array API standard and GPU computing. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- Added robust centered log-ratio transformation of compositional data. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- Calculates a distance matrix from a multiple sequence alignment. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)
- A new metric calculates evolutionary distance between aligned sequences. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)
- A new metric calculates evolutionary distance between aligned sequences. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)
- New plotting functionality includes centroids, confidence ellipses, and 2D plotting. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)
- A new format can read PHYLIP formatted distance matrices. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)

### Performance

- Improves performance of this differential abundance test. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Improves performance of this differential abundance test. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Expands Array API and GPU support, including Numba acceleration on NVIDIA and AMD GPUs. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds optional Numba acceleration for Mantel. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds optional Numba acceleration for PERMANOVA. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Adds optional Numba acceleration for PERMDISP. (0.7.4) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.4)
- Significantly accelerated the minimum evolution phylogenetics algorithm. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- Significantly accelerated the minimum evolution phylogenetics algorithm. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- Significantly reduced the time required to import scikit-bio. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- Significantly accelerated the minimum evolution phylogenetics algorithm. (0.7.3) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.3)
- Can now operate directly on condensed form `DistanceMatrix` objects, reducing memory footprint by 50%. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)
- Can now operate directly on condensed form `DistanceMatrix` objects, reducing memory footprint by 50%. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)
- The method is more efficient and can accept array-like input. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)

### Behavior and default changes

- The existing format now supports relaxed PHYLIP formatted multiple sequence alignments. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)

### Fixes and other changes

- Most of scikit-bio is compatible, but users who rely heavily on the metadata module are advised to pin Pandas below 3.0 until a later release. (0.7.2) — [release notes](https://github.com/scikit-bio/scikit-bio/releases/tag/0.7.2)