Advances in anti-aging and age-related disease drug discovery based on artificial intelligence and multi-omics

Aging is a highly complex biological process driven by the progressive accumulation of cellular damage, ultimately leading to functional decline and increased susceptibility to chronic diseases. As aging increases the risk of most chronic diseases, it involves changes across many biological systems that may not be adequately treated with the traditional “one disease-one target-one drug” strategy. These conventional approaches are not only time-consuming and costly but also difficult to capture the complex, interconnected biological changes that occur during aging. Recent advances in high-throughput multi-omics technologies, including genomics, transcriptomics, proteomics, epigenomics, and metabolomics, have enabled the study of aging from a systems perspective. By integrating these large-scale datasets, researchers have gained more profound insights into aging-related molecular pathways and regulatory networks. This progress has also enabled the development of aging clocks, such as Horvath, PhenoAge, and GrimAge, which use molecular features to estimate biological age and assess the effects of anti-aging interventions. In parallel, artificial intelligence (AI) has emerged as a transformative force in anti-aging drug discovery by enabling efficient integration, interpretation, and prediction across complex biological datasets. This review summarizes three major AI-driven strategies. First, AI-enabled drug repositioning leverages transcriptomic perturbation signatures, exemplified by platforms such as the Connectivity Map, to identify novel approved compounds against aging, including lifespan-extending natural products such as oridonin. Second, AI-powered virtual screening approaches that integrate molecular docking, machine learning, and graph neural networks have accelerated the identification of novel senolytics, including compounds with superior potency compared to first-generation agents. Third, generative AI models, such as variational autoencoders, generative adversarial networks, and Transformer-based architectures, have enabled the de novo design of drug-like molecules optimized for aging-related targets, as exemplified by the AI-assisted discovery of the TNIK inhibitor INS018₀55. Despite these advances, several challenges remain, including differences between omics datasets, limited interpretability of deep learning models, and difficulties in translating findings from animal models to humans. Future progress will require more interpretable AI methods, personalized multi-omics aging clocks, and greater use of human-related experimental systems and real-world clinical data. Together, the integration of AI and multi-omics technologies provides a strong foundation for precision anti-aging medicine discovery and offers new opportunities to advance translational aging research.