# SurfFold: a unified model for protein inverse folding by integrating surface and structural information

## TL;DR

SurfFold improves protein sequence prediction by combining backbone and surface data, leading to better accuracy in drug discovery applications.

## Contribution

SurfFold introduces a unified inverse folding model that integrates surface and structural information with a novel representation alignment module.

## Key findings

- SurfFold achieves state-of-the-art performance on the CATH4.2 dataset.
- The model effectively incorporates side-chain and surface interactions for sequence prediction.
- Homologous structure experiments show SurfFold's strong performance in protein design.

## Abstract

Proteins play a crucial role in biological systems, and accurate protein sequence prediction is essential for applications such as drug discovery. Existing inverse folding models primarily rely on protein backbone structure information, overlooking the biochemical properties embedded in protein surface data that constrain its functionality, leading to limited prediction accuracy.

We propose a novel inverse folding framework, SurfFold, which integrates both protein backbone structure and surface information for sequence prediction. Additionally, it incorporates side-chain structural information and its interaction with surface information. Then, we introduce a Representation alignment module to better fuze structure and surface Representations. Experimental results demonstrate that SurfFold achieves state-of-the-art performance on the CATH4.2 dataset, and additional experiments validate the effectiveness of the proposed modules. Moreover, the homologous structure inverse folding experiment also demonstrates that SurfFold possesses excellent capability in homologous protein design.

The source code and data are available at https://github.com/jiudizhengf/SurfFold.