[2603.01511] Multimodal Mixture-of-Experts with Retrieval Augmentation for Protein Active Site Identification

Computer Science > Artificial Intelligence arXiv:2603.01511 (cs) [Submitted on 2 Mar 2026] Title:Multimodal Mixture-of-Experts with Retrieval Augmentation for Protein Active Site Identification Authors:Jiayang Wu, Jiale Zhou, Xingyi Zhang, Xun Lin, Tianxu Lv, Leong Hou U, Rubo Wang, Yefeng Zheng View a PDF of the paper titled Multimodal Mixture-of-Experts with Retrieval Augmentation for Protein Active Site Identification, by Jiayang Wu and 7 other authors View PDF HTML (experimental) Abstract:Accurate identification of protein active sites at the residue level is crucial for understanding protein function and advancing drug discovery. However, current methods face two critical challenges: vulnerability in single-instance prediction due to sparse training data, and inadequate modality reliability estimation that leads to performance degradation when unreliable modalities dominate fusion processes. To address these challenges, we introduce Multimodal Mixture-of-Experts with Retrieval Augmentation (MERA), the first retrieval-augmented framework for protein active site identification. MERA employs hierarchical multi-expert retrieval that dynamically aggregates contextual information from chain, sequence, and active-site perspectives through residue-level mixture-of-experts gating. To prevent modality degradation, we propose a reliability-aware fusion strategy based on Dempster-Shafer evidence theory that quantifies modality trustworthiness through belief mass functions and lea...