RadDiff: Retrieval-Augmented Denoising Diffusion for Protein Inverse Folding

ArXi:2512.00126v2 Announce Type: replace-cross Protein inverse folding, the design of an amino acid sequence based on a target protein structure, is a fundamental problem of computational protein engineering. Existing methods either generate sequences without leveraging external knowledge or relying on protein language models~(PLMs). The former omits the knowledge d in natural protein data, while the latter is parameter-inefficient and inflexible to adapt to ever-growing protein data.