Deciphering Fingerprints of 3D Molecular Surfaces for Accurate Epitope Prediction

2026-06-22

AI Summary

This work addresses the limitations of existing epitope prediction methods, which predominantly rely on sequence or backbone structural information and struggle to accurately identify discontinuous epitopes determined by molecular surface properties. To overcome this challenge, the authors propose SurfBind, the first antibody–antigen interaction modeling framework centered explicitly on three-dimensional molecular surfaces. SurfBind integrates geometric and physicochemical features through a surface patch representation, a binding-state-aware cross-attention mechanism, and a coarse-to-fine hierarchical prediction strategy. Built upon a Transformer architecture, SurfBind achieves state-of-the-art performance on established benchmarks such as SAbDab and DB5.5, demonstrating strong generalization capabilities to unseen antibodies and diverse conformational states.

Abstract

Molecular surfaces encode the geometric and physicochemical patterns that determine antibody-antigen recognition, central to epitope prediction. However, existing methods rely on sequences or backbone structures and struggle to capture discontinuous, surface-driven epitopes. This study presents SurfBind, a surface-centric learning framework for epitope prediction that operates directly on molecular surface representations. SurfBind integrates geometric and physicochemical cues through a Transformer-based architecture with patch-level surface modeling, binder-aware cross-attention, and a hierarchical coarse-to-fine prediction paradigm. Experiments on challenging epitope identification benchmarks, including SAbDab and DB5.5, demonstrate that SurfBind achieves state-of-the-art performance and strong generalization across unseen antibodies and conformational states, highlighting the value of interaction-aware surface modeling for understanding the crucial mechanisms of protein-protein interactions.

Problem

• epitope prediction
• molecular surface
• antibody-antigen recognition
• discontinuous epitopes
• protein-protein interactions

Innovation

• molecular surface
• epitope prediction
• Transformer architecture
• surface-centric learning
• protein-protein interaction