DiffMeta-RL: Reinforcement Learning-Guided Graph Diffusion for Metabolically Stable Molecular Generation
Journal of Chemical Information and Modeling · 10.11.2025 · 65 (21) · 11746-11765
DOI: 10.1021/acs.jcim.5c02060 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Chemical Information and Modeling, online erschienen am 29.10.2025. Verfasst von Dayan Liu und 7 weiteren Autor:innen. Beteiligte Einrichtung: College of Computer Science and Technology, China University of Petroleum (East China), Qingdao 266580, Shandong, China. Hauptthemen laut MEDLINE: Cytochrome P-450 Enzyme System, Proton Pump Inhibitors. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 80 Literaturangaben. Gefördert von: National Key Research and Development Program of China, Taishan Scholar Project of Shandong Province, Fundamental Research Funds for the Central Universities.
Abstract (Originalsprache)
Modern drug discovery faces significant challenges in optimizing compounds for favorable ADMET properties, notably metabolic stability, as rapid metabolism impacts bioavailability, half-life, and drug–drug interactions. Central to these issues are cytochrome P450 (CYP450) enzymes, which metabolize most approved drugs, with unintended inhibition or accelerated clearance contributing substantially to clinical failures. To address this, we developed DiffMeta-RL, a discrete graph diffusion model enhanced with reinforcement learning, enabling controllable optimization of pharmacological properties. Unlike conventional actor-critic frameworks that require a separate policy network, DiffMeta-RL introduces a streamlined approach by directly optimizing the generative diffusion model itself as the policy, embedding reward-guided updates into the denoising process. Additionally, we proposed MetaCYP, a highly accurate CYP450 inhibition predictor, integrating it into DiffMeta-RL’s reward function to guide molecular generation toward reduced CYP450 liability and improved metabolic stability. In benchmark evaluations, DiffMeta-RL outperforms existing models in validity, uniqueness, and novelty. Specifically applied to designing proton pump inhibitors (PPIs) targeting gastric H + /K + -ATPase, DiffMeta-RL successfully generated candidates exhibiting favorable predicted binding affinity and reduced susceptibility to CYP2C19 and CYP3A4 metabolism, highlighting its practical potential for advancing drug discovery efforts.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- College of Computer Science and Technology, China University of Petroleum (East China), Qingdao 266580, Shandong, China.
- Shandong Key Laboratory of Intelligent Oil & Gas Industrial Software, Qingdao 266580, Shandong, China.
- Institute of Big Data, Central South University, Changsha 410083, PR China.
- The Interdisciplinary Graduate Program in Integrative Biotechnology and Translational Medicine, Yonsei University, Incheon 21983, Korea.
- State Key Laboratory of Chemical Safety, Qingdao 266580, Shandong, China.
ORCID-Profile: Dayan Liu, Tao Song, Shuang Wang, Xue Li
Förderung
- National Key Research and Development Program of China
- Taishan Scholar Project of Shandong Province
- Fundamental Research Funds for the Central Universities
- National Natural Science Foundation of China
- Natural Science Foundation of Shandong Province
- Foundation of Science and Technology Development of Jinan
Zitieren
APA
Liu, D., Song, T., Wang, S., Li, X., Han, P., Ju, G., Wang, J., & Wang, S. (2025). DiffMeta-RL: Reinforcement Learning-Guided Graph Diffusion for Metabolically Stable Molecular Generation. Journal of Chemical Information and Modeling, 65(21), 11746-11765. https://doi.org/10.1021/acs.jcim.5c02060
Vancouver
Liu D, Song T, Wang S, Li X, Han P, Ju G, et al. DiffMeta-RL: Reinforcement Learning-Guided Graph Diffusion for Metabolically Stable Molecular Generation. Journal of Chemical Information and Modeling. 2025;65(21):11746-11765. doi:10.1021/acs.jcim.5c02060
BibTeX
@article{Liu2025,
author = {Liu, Dayan and Song, Tao and Wang, Shuang and Li, Xue and Han, Peifu and Ju, Gehang and Wang, Jianmin and Wang, Shudong},
title = {DiffMeta-RL: Reinforcement Learning-Guided Graph Diffusion for Metabolically Stable Molecular Generation},
journal = {Journal of Chemical Information and Modeling},
year = {2025},
volume = {65},
number = {21},
pages = {11746--11765},
doi = {10.1021/acs.jcim.5c02060}
}
EUVIMED