FRAIL: fragment-based reinforcement learning for molecular design and benchmarking on fatty acid amide hydrolase 1 (FAAH-1)

Luong, Manh-Tu ; Pham, Khanh Huyen Thi ; Nguyen, Nhat-Hai ; Le, Van-Tuan ; Pham, Phu Tran Vinh ; Nguyen, Tan Khanh ; Nguyen, Thi-Thu (Universidad de Zaragoza)

Resumen: We propose FRAIL (Fragment-based Reinforcement Learning for Inhibitors), a generative AI framework that integrates fragment-based molecular design, multi- objective reinforcement learning, and molecular modeling to accelerate inhibitor discovery. Several deep generative models were fine-tuned on FAAH-1 (Fatty Acid Amide Hydrolase 1)–specific dataset and systematically benchmarked, with the best-performing model incorporated into FRAIL. The framework employs a customized reward function that jointly optimizes physicochemical properties and predicted bioactivity (pIC50) to guide molecular generation toward FAAH- favorable chemotypes. FRAIL generated structurally novel, fragment-grown compounds exhibiting high predicted binding affinity, desirable drug-likeness, and synthetic accessibility. These findings demonstrate FRAIL’s capability to enhance rational drug design and provide a reproducible pipeline for the discovery of experimentally viable FAAH inhibitors.

Idioma: Inglés

DOI: 10.1007/s11030-025-11448-4

Año: 2026

Publicado en: MOLECULAR DIVERSITY
ISSN: 1381-1991

Tipo y forma: Article (Published version)

Área (Departamento): Área Bioquímica y Biolog.Mole. (Dpto. Bioq.Biolog.Mol. Celular)