Multi-Modality Representation Learning for Antibody-Antigen Interactions Prediction

2025-03-22

AI Summary

Antibody–antigen interaction (AAI) prediction is hindered by the scarcity of sequence–structure paired data and limited generalization due to existing methods’ neglect of fine-grained 3D antibody structural features and cross-antibody sequence similarities. To address this, we propose the first multimodal graph learning framework integrating 1D sequences and 3D structures: it constructs a hierarchical antibody graph representation capturing intra-residue conformational relationships and inter-antibody sequence homologies; introduces a hierarchical graph representation learning module synergizing normalized adaptive graph convolutional networks (NAGCN) and graph attention networks (GAT); and releases AAI-Bench—the first benchmark dataset featuring experimentally validated structures, sequences, and interaction labels. Our method achieves significant performance gains over state-of-the-art approaches on AAI-Bench. The code and dataset are publicly available.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Humans and AI: Human-AI Collaboration / Human-AI Teaming

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

While deep learning models play a crucial role in predicting antibody-antigen interactions (AAI), the scarcity of publicly available sequence-structure pairings constrains their generalization. Current AAI methods often focus on residue-level static details, overlooking fine-grained structural representations of antibodies and their inter-antibody similarities. To tackle this challenge, we introduce a multi-modality representation approach that integates 3D structural and 1D sequence data to unravel intricate intra-antibody hierarchical relationships. By harnessing these representations, we present MuLAAIP, an AAI prediction framework that utilizes graph attention networks to illuminate graph-level structural features and normalized adaptive graph convolution networks to capture inter-antibody sequence associations. Furthermore, we have curated an AAI benchmark dataset comprising both structural and sequence information along with interaction labels. Through extensive experiments on this benchmark, our results demonstrate that MuLAAIP outperforms current state-of-the-art methods in terms of predictive performance. The implementation code and dataset are publicly available at https://github.com/trashTian/MuLAAIP for reproducibility.

Problem

• Predict antibody-antigen interactions with limited sequence-structure data
• Overcome lack of fine-grained structural antibody representations
• Integrate 3D and 1D data to model intra-antibody relationships

Innovation

• Multi-modality representation combining 3D and 1D data
• Graph attention networks for structural features
• Normalized adaptive graph convolution for sequence associations