Refold: Refining Protein Inverse Folding with Efficient Structural Matching and Fusion

TL;DR

Refold is a new framework that combines database-derived structural priors with deep learning to improve protein inverse folding, achieving state-of-the-art accuracy and better handling of uncertain regions.

Contribution

Refold introduces a synergistic method that integrates structural priors with deep learning predictions, including a Dynamic Utility Gate to manage prior quality, advancing protein inverse folding.

Findings

Achieves 0.63 native sequence recovery on benchmarks.

Delivers larger improvements in high-uncertainty regions.

Outperforms existing methods in standard evaluations.

Abstract

Protein inverse folding aims to design an amino acid sequence that will fold into a given backbone structure, serving as a central task in protein design. Two main paradigms have been widely explored. Template-based methods exploit database-derived structural priors and can achieve high local precision when close structural neighbors are available, but their dependence on database coverage and match quality often degrades performance on out-of-distribution (OOD) targets. Deep learning approaches, in contrast, learn general structure-to-sequence regularities and usually generalize better to new backbones. However, they struggle to capture fine-grained local structure, which can cause uncertain residue predictions and missed local motifs in ambiguous regions. We introduce Refold, a novel framework that synergistically integrates the strengths of database-derived structural priors and deep learning prediction to enhance inverse folding. Refold obtains structural priors from matched neighbors and fuses them with model predictions to refine residue probabilities. In practice, low-quality neighbors can introduce noise, potentially degrading model performance. We address this issue with a Dynamic Utility Gate that controls prior injection and falls back to the base prediction when the priors are untrustworthy. Comprehensive evaluations on standard benchmarks demonstrate that Refold achieves state-of-the-art native sequence recovery of 0.63 on both CATH 4.2 and CATH 4.3. Also, analysis indicates that Refold delivers larger gains on high-uncertainty regions, reflecting the complementarity between structural priors and deep learning predictions.