Construction of a Multimodal Ageing Foundation Model for Ageing Mechanism Elucidation and Intervention Discovery

This project aims to develop a multimodal aging foundation model by integrating electronic health records, medical imaging, lifestyle factors, biomarkers, multi-omics profiles, and longitudinal outcomes. The model will learn a unified representation of aging to jointly characterize biological age, organ-specific aging, frailty, and metabolic health. We will further construct an intervention knowledge graph and prioritization module to identify candidate anti-aging interventions, followed by validation in mouse and dog models. This study will establish a closed-loop framework integrating population-scale modeling, intervention discovery, animal validation, and iterative model optimization for healthy aging research.