Welcome to the GitHub repository for Inferring selection effects in SARS-CoV-2 with Bayesian Viral Allele Selection. The results from our main analysis are available in paper. More recent inference results are also available: 08.10.22, 09.19.22, 05.15.23, 06.13.23, 06.26.23, 09.04.23, 09.18.23, 10.30.23, 11.13.23, 11.27.23, 12.11.23, 01.01.24, 01.22.24, 02.05.24, 02.22.24, 03.04.24, 03.22.24, 04.04.24, 05.27.24, 06.24.24, and 07.22.24. Thanks to Benjamin Kotzen (@bkotzen) for help generating these more recent results!
BVAS requires Python 3.8 or later and the following Python packages: PyTorch, pandas, and pyro.
Note that if you wish to run BVAS on a GPU you need to install PyTorch with CUDA support. In particular if you run the following command from your terminal it should report True:
python -c 'import torch; print(torch.cuda.is_available())'
Install directly from GitHub:
pip install git+https://github.com/broadinstitute/bvas.git
Install from source:
git clone git@github.com:broadinstitute/bvas.git
cd bvas
pip install .
The documentation is available here.
This repo is organized as follows:
- bvas: all the core code: inference algorithms and simulations
- paper: some of the figures and inference results contained in the paper
- notebooks: Jupyter notebooks demonstrating BVAS usage
- basic_demo.ipynb: demo using simulated data
- S_gene_demo.ipynb: demo using GISAID data restricted only to the SARS-CoV-2 S gene
- data: pre-processing scripts and (some of the) data used in the analysis
- docs: source code for the documentation
- example_scripts: example scripts that demo BVAS usage
- tests: unit tests for verifying the correctness of inference algorithms and other code