SAWNERGY: A Python framework for dynamic residue-interaction networks and walk-based embeddings from molecular dynamics simulations

SAWNERGY is a modular Python toolkit that converts MD trajectories into residue interaction networks (RINs), samples random and self-avoiding walks (including time-aware variants), and trains per-frame skip-gram embeddings using PureML or PyTorch backends. Artifacts are stored as compressed Zarr archives with rich metadata for reproducibility and downstream analysis.