README.md

scANNA (Package Repository)

This repository hosts the package for scANNA: single-cell ANalysis using Neural Attention (Preprint, submitted). To make package maintenance more efficient, and to provide more specific tutorials on using scANNA, we have located tutorials in a dedicated repository, as listed below.

Installing scANNA

Installing the GitHub Repository (Recommended)

scANNA requires Python 3.10 and can be installed using PyPI:

$ pip install git+https://github.com/SindiLab/scANNA.git

or can be first cloned and then installed as the following:

$ git clone https://github.com/SindiLab/scANNA.git
$ pip install ./scANNA

Install Package Locally with pip

Once the files are available, make sure to be in the same directory as setup.py. Then, using pip, run:

pip install -e .

In the case that you want to install the requirements explicitly, you can do so by:

pip install -r requirements.txt

Although the core requirements are listed directly in setup.py. Nonetheless, it is good to run this beforehand in case of any dependecies conflicts.

Training scANNA

All main scripts for training (and finetuning) our deep learning model are located in the training_and_finetuning_scripts folder in this repository.

Tutorials

We have compiled a set of notebooks as tutorials to showcase scANNA's capabilities and interptretability. These notebooks located here.

Please feel free to open issues for any questions or requests for additional tutorials!

Trained Models

TODO: Will be released with the next preprint for scANNA.