moscot 0.1.0

Multi-omic single-cell optimal transport tools

moscot - multi-omic single-cell optimal transport tools

moscot is a general framework to apply tools from optimal transport to time-course single-cell data. It supports:

• single-cell RNA-seq and ATAC-seq data (paired and unpaired)

• single-cell lineage-traced data (prospective and retrospective)

while scaling to large cell numbers. In the backend, moscot is powered by OTT which is a Jax-based optimal transport toolkit that supports just-in-time compilation, automatic differentiation and linear memory complexity for OT problems.

Installation

In order to install moscot, run:

git clone https://github.com/theislab/moscot
cd moscot
pip install -e.'[dev]'
pre-commit install

for pre-commit you might have to install prettier with conda: conda install -c conda-forge prettier.

If used with GPU, additionally run:

pip install --upgrade "jax[cuda]" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html