Artificial intelligence-assisted discovery of natural product geroprotectors: target prediction, aging biomarker identification, and drug repurposing.

Abstract

Natural products derived from traditional medicines are valuable sources of potential geroprotective compounds; however, their translation into therapeutic candidates is limited by chemical complexity, polypharmacology, limited mechanistic resolution, and lengthy experimental validation. Artificial intelligence (AI) offers an emerging framework for integrating chemical, pharmacological, multi-omics, imaging, and literature derived data to accelerate natural product anti-aging research. This review summarizes recent applications of AI in three interconnected areas: drug repurposing, prediction of aging-related targets and signaling pathways, and identification of biomarkers for biological aging and cellular senescence. AI integrates chemical, molecular, omics, and imaging data to prioritize candidate compounds, predict drug-target interactions, identify synergistic combinations, and reveal mechanisms underlying geroprotection. However, computational predictions should not be interpreted as evidence of efficacy or causality. Data heterogeneity, sampling bias, limited model interpretability, inadequate external validation, and insufficient pharmacokinetic and long-term safety assessment remain major barriers to translation. Future progress will require an iterative framework that combines AI prediction with target engagement studies, senescence models, organoids, animal experiments, and clinical biomarker validation. This integrated strategy may improve the efficiency, reproducibility, and translational potential of natural product geroscience research.