This project aims to decode the biological continuum linking cardiovascular and cerebrovascular diseases (CVD-CeVD) through an integrated multiomic and artificial intelligence (AI) framework. Coronary heart disease (CHD) and stroke, though traditionally treated as separate entities, share overlapping genetic, metabolic, and inflammatory origins that reflect dysfunction of the “cardio-cerebral axis.”
Research questions and objectives:
We will investigate how molecular, metabolic, and imaging networks converge or diverge across CHD and stroke. Specific objectives include:
(1) mapping shared molecular signatures along the cardio-cerebral axis;
(2) identifying subgroup-specific mechanisms across age and sex strata;
(3) constructing AI-based integrative models to predict comorbid trajectories;
(4) exploring causal pathways linking omic alterations to clinical outcomes.
Scientific rationale:
Traditional studies have treated cardiovascular and cerebrovascular diseases independently, overlooking their biological interplay. Our preliminary analyses suggest overlapping metabolic-inflammatory mechanisms, yet their dynamic evolution remains unclear. This study will integrate genomics, epigenomics, proteomics, metabolomics, imaging, and clinical data within a unified framework. Graph neural networks and transformer-based architectures will capture cross-modal relationships and temporal dynamics, while causal inference and explainable AI will elucidate biological drivers of disease progression. Leveraging UK Biobank’s multimodal depth and population scale, this project will redefine cardiovascular-cerebrovascular comorbidity as a dynamic, data-driven continuum, providing mechanistic insights and translational targets for precision medicine.
