Antibody Design and Optimization with Multi-scale Equivariant Graph Diffusion Models for Accurate Complex Antigen Binding

TL;DR

This paper introduces AbMEGD, a novel geometric deep learning framework using equivariant graph diffusion for antibody design, significantly improving accuracy and generalization in complex antigen binding predictions.

Contribution

AbMEGD is the first end-to-end multi-scale equivariant graph diffusion model for antibody sequence and structure co-design, enhancing geometric fidelity and generalization over prior methods.

Findings

10.13% increase in amino acid recovery

3.32% improvement percentage rise

0.062 Å reduction in RMSD in CDR-H3 region

Abstract

Antibody design remains a critical challenge in therapeutic and diagnostic development, particularly for complex antigens with diverse binding interfaces. Current computational methods face two main limitations: (1) capturing geometric features while preserving symmetries, and (2) generalizing novel antigen interfaces. Despite recent advancements, these methods often fail to accurately capture molecular interactions and maintain structural integrity. To address these challenges, we propose \textbf{AbMEGD}, an end-to-end framework integrating \textbf{M}ulti-scale \textbf{E}quivariant \textbf{G}raph \textbf{D}iffusion for antibody sequence and structure co-design. Leveraging advanced geometric deep learning, AbMEGD combines atomic-level geometric features with residue-level embeddings, capturing local atomic details and global sequence-structure interactions. Its E(3)-equivariant diffusion method ensures geometric precision, computational efficiency, and robust generalizability for complex antigens. Furthermore, experiments using the SAbDab database demonstrate a 10.13\% increase in amino acid recovery, 3.32\% rise in improvement percentage, and a 0.062~\AA\ reduction in root mean square deviation within the critical CDR-H3 region compared to DiffAb, a leading antibody design model. These results highlight AbMEGD's ability to balance structural integrity with improved functionality, establishing a new benchmark for sequence-structure co-design and affinity optimization. The code is available at: https://github.com/Patrick221215/AbMEGD.