Sovereign 8s.is mirror of standardgalactic/bascvi
  • Python 58.7%
  • Jupyter Notebook 41.3%
Find a file
Repository files (latest commit first)
Filename Latest commit message Latest commit date
2025-11-25 10:28:55 +00:00
.github/ISSUE_TEMPLATE Update issue templates 2024-09-16 16:18:09 -04:00
notebooks Update locally_embed_h5ad.ipynb 2025-09-12 14:22:08 +01:00
src atlas-building opt 2025-11-25 10:28:55 +00:00
.gitignore Merge branch 'main' of https://github.com/PhenomicAI/bascvi into macrogenes 2025-03-17 19:50:09 +00:00
pyproject.toml moved all source code into src directory 2025-02-05 13:49:15 +00:00
README.md Update README.md 2025-09-12 10:46:53 +01:00

BAscVI

This repository accompanies the paper:

“Adversarial learning enables unbiased organism-wide cross-species alignment of single-cell RNA data at scale”
arXiv preprint


Repository Overview

This repo contains three Python modules that can be installed separately depending on your use case:

  • pai
    API for inference of cell-type embeddings and cell-type labels from scRNA-seq datasets.
    Supports .h5ad formats (downloadable from cellxgene).

  • pai_soma_data
    A wrapper around TileDB for exploring the scREF atlas, used in our public notebooks.

  • ml_benchmarking
    Code to run and evaluate ML models on the scREF and scREF-mu benchmarks.


Installation

Clone the repository:

git clone https://github.com/PhenomicAI/bascvi.git
cd bascvi

pip install pai - recomended

pip install -U phenomic-ai  

Install pai from local

TODO: update github version to 1.11

pip install -e src/pai

Verify:

from pai.utils.option_choices import tissue_organ_option_choices
from pai.embed import PaiEmbeddings

Install pai_soma_data

pip install -e src/pai_soma_data

Verify:

from pai_soma_data import pai_soma_data

Install ml_benchmarking

pip install -e src/ml_benchmarking

Verify:

import ml_benchmarking.bascvi as bascvi

You can install any combination of these modules depending on your needs.


Example Usage

from pai.utils.option_choices import tissue_organ_option_choices
from pai.embed import PaiEmbeddings
from pai_soma_data import pai_soma_data
import ml_benchmarking.bascvi as bascvi

License

Everything is released under an MIT license. Please feel free to use it, but cite us as we have cited others.