CAME-AB: Cross-Modality Attention with Mixture-of-Experts for Antibody Binding Site Prediction

2025-09-08

AI Summary

Existing antibody binding site prediction methods rely on unimodal feature representations, limiting their ability to accurately identify antigen regions specifically recognized by antibodies. To address this, we propose MAbBind, a cross-modal attention framework that integrates five complementary biological modalities: amino acid sequences, BLOSUM profiles, pretrained embeddings, structure-aware features, and GCN-optimized biochemical graphs. We introduce an adaptive multimodal fusion mechanism coupled with a supervised contrastive learning objective to dynamically weight modality contributions, thereby enhancing intra-class compactness and inter-class separability in the representation space. Furthermore, we adopt a Transformer-Mixture-of-Experts (MoE) architecture to improve modeling capacity. On multiple benchmark datasets, MAbBind achieves significant improvements in F1-score and AUC-ROC over state-of-the-art methods. Ablation studies confirm the effectiveness of both multimodal integration and individual components.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Antibody binding site prediction plays a pivotal role in computational immunology and therapeutic antibody design. Existing sequence or structure methods rely on single-view features and fail to identify antibody-specific binding sites on the antigens-a dual limitation in representation and prediction. In this paper, we propose CAME-AB, a novel Cross-modality Attention framework with a Mixture-of-Experts (MoE) backbone for robust antibody binding site prediction. CAME-AB integrates five biologically grounded modalities, including raw amino acid encodings, BLOSUM substitution profiles, pretrained language model embeddings, structure-aware features, and GCN-refined biochemical graphs-into a unified multimodal representation. To enhance adaptive cross-modal reasoning, we propose an adaptive modality fusion module that learns to dynamically weight each modality based on its global relevance and input-specific contribution. A Transformer encoder combined with an MoE module further promotes feature specialization and capacity expansion. We additionally incorporate a supervised contrastive learning objective to explicitly shape the latent space geometry, encouraging intra-class compactness and inter-class separability. To improve optimization stability and generalization, we apply stochastic weight averaging during training. Extensive experiments on benchmark antibody-antigen datasets demonstrate that CAME-AB consistently outperforms strong baselines on multiple metrics, including Precision, Recall, F1-score, AUC-ROC, and MCC. Ablation studies further validate the effectiveness of each architectural component and the benefit of multimodal feature integration. The model implementation details and the codes are available on https://anonymous.4open.science/r/CAME-AB-C525

Problem

• Predicting antibody binding sites on antigens
• Integrating multimodal biological features for representation
• Overcoming limitations of single-view sequence or structure methods

Innovation

• Integrates five biological modalities into unified representation
• Uses adaptive modality fusion for dynamic weighting
• Combines Transformer with MoE for feature specialization