A toolkit designed to perform some common tasks in protein engineering, especially encoding protein sequences (in various formats), assisting in fitting uncertainty-aware models to data, and using a trained model to search for improved candidates using the RESP pipeline from Parkinson et al. 2023 (Nature Communications).
For an example of the full RESP pipeline, including uncertainty-aware model training, please see this repo.
pip install resp_protein_toolkit
Alternatively, to install from github, clone this repo, then from inside the resulting directory run:
pip install . --verbose
For usage and general guidelines, see the docs.
If using this toolkit in work intended for publication, please cite:
