EUVIMEDEUVIMED entdecken
Crossref

GATRsite: RNA–Ligand Binding Site Prediction Using Graph Attention Networks and Pretrained RNA Language Models

Chuance Sun iD, Linghao Zhang, Lingfeng Zhang, Yuehua Song, Buyong Ma iD, Yanjing Wang iD

Journal of Chemical Information and Modeling · 25.08.2025 · 65 (16) · 8448-8461

DOI: 10.1021/acs.jcim.5c00605 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Chemical Information and Modeling, online erschienen am 15.08.2025. Verfasst von Chuance Sun und 5 weiteren Autor:innen. Beteiligte Einrichtung: Engineering Research Center of Cell & Therapeutic Antibody (MOE), School of Pharmacy, Shanghai Jiao Tong University, Shanghai 200240, China. Hauptthemen laut MEDLINE: RNA, Computational Biology. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 50 Literaturangaben. Gefördert von: Science and Technology Commission of Shanghai Municipality, National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Identifying functional sites of RNA, particularly those where small molecules bind, is crucial for understanding related biological processes and advancing drug design. Small molecule therapies, compared to traditional protein-targeted therapies, have the potential to pioneer novel RNA-specific therapeutic strategies. However, the challenge lies in developing accurate and efficient computational methods, requiring novel computational models that can better characterize RNA and precisely predict RNA-small molecule binding sites. In this study, we introduced GATRsite, an efficient deep learning framework leveraging graph attention networks (GATs) and Pretrained RNA Language Models to predict RNA-ligand binding sites. GATRsite regards RNA nucleotides as nodes, and its main component is an RNA graph with nodes that comprehensively incorporates both sequential and structural features. Furthermore, it integrates embeddings derived from advanced Pretrained RNA Language Models, which precisely capture the intricate structural and functional complexities of RNA molecules. GATRsite outperforms other state-of-the-art methods, particularly in terms of recall rates, Matthew's correlation coefficient, and F1 score on benchmark test sets. Moreover, GATRsite exhibits significant robustness regarding the predicted RNA structures. A user-friendly online server for GATRsite is freely available at https://malab.sjtu.edu.cn/GATRsite/.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

RNAComputational BiologyLigandsBinding SitesDeep LearningModels, Molecular

Substanzen

RNALigands

Autor:innen und Institutionen

  • Engineering Research Center of Cell & Therapeutic Antibody (MOE), School of Pharmacy, Shanghai Jiao Tong University, Shanghai 200240, China.
  • School of Electrical Engineering and Computer Science, University of Ottawa, 75 Laurier Avenue, Ottawa, Ontario K1N 6N5, Canada.
  • Faculty of Science, University of British Columbia, Kelowna, 3333 University Way, Kelowna, BC V1V 1V7, Canada.
  • Shanghai Digiwiser BioTechnolgy, Limited, Shanghai 201203, China.

ORCID-Profile: Chuance Sun, Buyong Ma, Yanjing Wang

Förderung

  • Science and Technology Commission of Shanghai Municipality
  • National Natural Science Foundation of China

Zitieren

APA

Sun, C., Zhang, L., Zhang, L., Song, Y., Ma, B., & Wang, Y. (2025). GATRsite: RNA–Ligand Binding Site Prediction Using Graph Attention Networks and Pretrained RNA Language Models. Journal of Chemical Information and Modeling, 65(16), 8448-8461. https://doi.org/10.1021/acs.jcim.5c00605

Vancouver

Sun C, Zhang L, Zhang L, Song Y, Ma B, Wang Y. GATRsite: RNA–Ligand Binding Site Prediction Using Graph Attention Networks and Pretrained RNA Language Models. Journal of Chemical Information and Modeling. 2025;65(16):8448-8461. doi:10.1021/acs.jcim.5c00605

BibTeX
@article{Sun2025,
  author = {Sun, Chuance and Zhang, Linghao and Zhang, Lingfeng and Song, Yuehua and Ma, Buyong and Wang, Yanjing},
  title = {GATRsite: RNA–Ligand Binding Site Prediction Using Graph Attention Networks and Pretrained RNA Language Models},
  journal = {Journal of Chemical Information and Modeling},
  year = {2025},
  volume = {65},
  number = {16},
  pages = {8448--8461},
  doi = {10.1021/acs.jcim.5c00605}
}
Kostenlos starten →