NesT-NABind: a Nested Transformer for Nucleic Acid-Binding Site Prediction on Protein Surface

• DOI: 10.1021/acs.jcim.4c01765
• Publication Date: 2025-01-17T13:32:10Z
• Authors: Xinyue Ma; Fenglei Li; Qianyu Chen; Shenghua Gao; Fang Bai
• Keywords: Surface protein

Protein-nucleic acid interactions play a crucial role in many physiological processes. Identifying the binding sites of nucleotides on protein surface is prerequisite for understanding molecular recognition mechanisms between two types macromolecules and also provides information to design or generate molecule modulators against these manipulate biological function according specific requirements. Existing studies mainly focus characterizing local surfaces around sites, often neglecting interrelationships among global information. To address this gap, we propose NesT-NABind, Nested Transformer Nucleic Acid-Binding site prediction. This model leverages Transformer's advanced capabilities contextual long-range dependency capturing. Specifically, introduce patch-scale process each protein-scale transformer integrate sequence entire protein. These Transformers operate at different scales protein, hence term "nested". Experiments demonstrate that NesT-NABind achieves 5.57% improvement F1 score 3.64% AUPRC compared state-of-the-art methods. With incorporation features, shows an enhanced predictive capability challenging large proteins therefore can be used much wider range applications.