This project aims to advance precision prevention of gastric disorders and their systemic complications by integrating lifestyle, genomics, and metabolomics data. Gastric diseases-such as gastritis, peptic ulcer, and gastric cancer-are increasingly recognized not only as local gastrointestinal conditions but also as contributors to cardio-metabolic dysfunction and cognitive decline. However, current risk prediction models rely heavily on clinical factors and lack the resolution needed for early, individualized intervention.
Our central research questions are: (1) How can integration of lifestyle, genomics, and metabolomics data improve early detection and risk stratification of gastric diseases compared to traditional clinical models? (2) Which specific combinations of genetic, metabolic, microbial, and lifestyle factors most strongly predict gastric disease onset, progression, and associated cardio-metabolic and cognitive complications? (3) How do metabolic and microbiome profiles mediate the relationship between gastric disease and cognitive dysfunction? (4) Can predictive models derived from multi-modal data identify individuals who would benefit most from personalized prevention strategies, such as dietary, microbial, or lifestyle interventions?
The scientific rationale of this project can be illustrated as follows: using data from the UK Biobank-including genomic, metabolomic, metagenomic, lifestyle, clinical, and cognitive assessments–we will apply machine learning and large language model-enhanced multi-agent systems to uncover complex inter-omics interactions and generate patient-centered insights. This integrative approach addresses the multifactorial nature of gastric disease and its extra-digestive manifestations, with the ultimate goal of enabling earlier diagnosis, targeted interventions, and improved long-term health outcomes across the cardio-metabolic-cognitive axis.
