Interpretable Protein−Nucleic Acid Binding Site Prediction via Domain-Guided Multimodal Learning
ACS Synthetic Biology · 2026
DOI: 10.1021/acssynbio.6c00340 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACS Synthetic Biology, 2026. Verfasst von Wenyu Xi und 4 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, University of Tsukuba, Tsukuba, Ibaraki305-8573, Japan. Hauptthemen laut MEDLINE: Deep Learning, RNA, DNA, Computational Biology. Die Arbeit nennt 60 Literaturangaben. Gefördert von: Japan Society for the Promotion of Science, Internal Research Grants of Macao Polytechnic University, Japan Science and Technology Agency.
Abstract (Originalsprache)
Accurate identification of protein−nucleic acid binding sites is essential for elucidating molecular mechanisms and for enabling structure-based drug discovery. Existing computational approaches struggle to jointly model sequence, structure, and domain knowledge, which limits both predictive accuracy and interpretability. We present BiMTBind, a domain-guided multimodal deep learning framework that integrates bidirectional state-space modeling with Transformer-based attention for residue-level binding site prediction. The model explicitly incorporates protein-domain annotations as biological priors to guide attention toward functionally relevant regions. Multi-source featurescontextual embeddings from protein language models and handcrafted structural descriptorsare fused via a learnable gating mechanism that adaptively weights each modality. This design enables the model to capture both global dependencies and local context. Extensive evaluations across diverse DNA and RNA benchmarks show that BiMTBind consistently outperforms state-of-the-art sequence- and structure-based methods on multiple metrics, including F1, MCC, and AUC. Case studies further show that domain-guided attention improves interpretability by aligning predicted binding residues with annotated functional domains. These results highlight the effectiveness of combining biologically informed priors with deep multimodal representation learning, offering a scalable and generalizable approach for protein−nucleic acid binding site prediction.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Computer Science, University of Tsukuba, Tsukuba, Ibaraki305-8573, Japan.
- University of Texas Southwestern Medical Center , Dallas, Texas75390, United States.
- Centre for Artificial Intelligence Driven Drug Discovery, Faculty of Applied Sciences, Macao Polytechnic University, Macao999078, China.
ORCID-Profile: Wenyu Xi, Tetsuya Sakurai, Leyi Wei
Förderung
- Japan Society for the Promotion of Science
- Internal Research Grants of Macao Polytechnic University
- Japan Science and Technology Agency
Zitieren
APA
Xi, W., Wang, R., Ye, X., Sakurai, T., & Wei, L. (2026). Interpretable Protein−Nucleic Acid Binding Site Prediction via Domain-Guided Multimodal Learning. ACS Synthetic Biology. https://doi.org/10.1021/acssynbio.6c00340
Vancouver
Xi W, Wang R, Ye X, Sakurai T, Wei L. Interpretable Protein−Nucleic Acid Binding Site Prediction via Domain-Guided Multimodal Learning. ACS Synthetic Biology. 2026. doi:10.1021/acssynbio.6c00340
BibTeX
@article{Xi2026,
author = {Xi, Wenyu and Wang, Ruheng and Ye, Xiucai and Sakurai, Tetsuya and Wei, Leyi},
title = {Interpretable Protein−Nucleic Acid Binding Site Prediction via Domain-Guided Multimodal Learning},
journal = {ACS Synthetic Biology},
year = {2026},
doi = {10.1021/acssynbio.6c00340}
}
EUVIMED