EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Leveraging 3D Molecular Spatial Visual Information and Multi‐Perspective Representations for Drug Discovery

Zimai Zhang iD, Xi Zhou, Yujie Qi, Xiaobo Zhu, Xun Deng, Feng Tan, Yuan Huang, Lun Hu, Zhuhong You, Pengwei Hu iD

Advanced Science · 2026-01 · 13 (2)

DOI: 10.1002/advs.202512453 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advanced Science, online erschienen am 15.10.2025. Verfasst von Zimai Zhang und 9 weiteren Autor:innen. Beteiligte Einrichtung: Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Urumqi, 830011, China. Hauptthemen laut MEDLINE: Drug Discovery, Deep Learning. Die Arbeit wurde bisher 37-mal zitiert (Crossref) und nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Xinjiang Uygur Autonomous Region, Xinjiang Tianchi Talent Program.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Drug discovery remains a costly and time‐intensive process, where accurate identification of drug associations is critical for therapeutic development. Existing computational approaches predominantly rely on sequence‐derived or 2D molecular representations, often overlooking the intrinsic 3D complexity of small molecules. Here, a deep learning framework is presented that directly learns from 3D molecular spatial visual information, capturing geometric, topological, and stereochemical features from spatial renderings. By integrating this spatial information with traditional molecular descriptors, unified multi‐perspective representations are constructed that better reflect molecular structure and function. Across benchmark tasks involving drug–microRNA, drug–drug, and drug–protein interaction prediction, this model consistently outperforms conventional fingerprint‐based baselines. Interpretability analyses show that the model attends to biologically relevant substructures, highlighting the value of 3D molecular spatial visual information in molecular recognition. These findings demonstrate the potential of spatially informed learning to enhance predictive performance and provide mechanistic insights in computational drug discovery.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Drug DiscoveryDeep LearningComputational BiologyHumansMicroRNAs

Stichwörter der Autor:innen

drug 3D molecular spatial visual informationdrug discoverymulti‐perspective learning

Substanzen

MicroRNAs

Autor:innen und Institutionen

  • Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Urumqi, 830011, China.
  • School of Software, Xinjiang University, Urumqi, 830046, China.
  • School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing, 100049, China.
  • Al and Quantum Lab, Merck KGaA, 64293, Darmstadt, Germany.
  • School of Computer Science, Northwestern Polytechnical University, Xian, 710129, China.

ORCID-Profile: Zimai Zhang, Pengwei Hu

Förderung

  • National Natural Science Foundation of China
  • Natural Science Foundation of Xinjiang Uygur Autonomous Region
  • Xinjiang Tianchi Talent Program
  • Tianshan Talent Training Program

Zitieren

APA

Zhang, Z., Zhou, X., Qi, Y., Zhu, X., Deng, X., Tan, F., Huang, Y., Hu, L., You, Z., & Hu, P. (2026). Leveraging 3D Molecular Spatial Visual Information and Multi‐Perspective Representations for Drug Discovery. Advanced Science, 13(2). https://doi.org/10.1002/advs.202512453

Vancouver

Zhang Z, Zhou X, Qi Y, Zhu X, Deng X, Tan F, et al. Leveraging 3D Molecular Spatial Visual Information and Multi‐Perspective Representations for Drug Discovery. Advanced Science. 2026;13(2). doi:10.1002/advs.202512453

BibTeX
@article{Zhang2026,
  author = {Zhang, Zimai and Zhou, Xi and Qi, Yujie and Zhu, Xiaobo and Deng, Xun and Tan, Feng and Huang, Yuan and Hu, Lun and You, Zhuhong and Hu, Pengwei},
  title = {Leveraging 3D Molecular Spatial Visual Information and Multi‐Perspective Representations for Drug Discovery},
  journal = {Advanced Science},
  year = {2026},
  volume = {13},
  number = {2},
  doi = {10.1002/advs.202512453}
}
Kostenlos starten →