Repository navigation
Releases
Releases · kumarshreshtha/protuna
v0.1.0 — Initial release
First public release of Protuna, an optimization framework for AI-guided protein engineering campaigns.
Features
- Modular optimization pipelines with configurable samplers, predictors, constraints, and acquisition strategies.
- Guided sequence generation using Feynman–Kac steering, classifier guidance, and beam search.
- Ray-based distributed inference for scalable screening.
- Protein scoring functions and access to preprocessed benchmark datasets.
Installation
Requires Python 3.12 or higher.
pip install protuna
Documentation
See https://protuna.readthedocs.io/en/latest/ for quickstarts, concepts, and guides