Research Question and Objective
Gastrointestinal disorders-including colorectal cancer, inflammatory bowel disease, non-alcoholic fatty liver disease, and peptic ulcer-are escalating global health threats. Although genetic and lifestyle determinants are established, how environmental exposures, psychological stress, and multi-omic signatures jointly orchestrate disease initiation and progression remains elusive. Harnessing UK Biobank’s deep phenotyping, we will (1) discover novel biomarkers via integrated omics-environment-psychosocial analyses; (2) dissect causal chains linking modifiable exposures, molecular perturbations, and clinical trajectories; and (3) engineer machine-learning algorithms for early risk prediction and precision prevention.
Scientific Rationale
Colorectal cancer dominates gastrointestinal malignancy mortality, driven by inherited risk, metabolic dysfunction, and lifestyle. Radiotherapy for pelvic cancers frequently precipitates chronic radiation enteropathy; despite differing triggers, it mirrors sporadic disorders through sustained inflammation and immune dysregulation, providing a tractable cross-disease paradigm. Circulating proteomes, metabolomes, and quantitative imaging phenotypes illuminate upstream pathways. Air pollution, ultra-processed diets, and sedentary behaviour further perturb the gut-brain axis by reshaping microbiota, immunity, and neuroendocrine signalling. Existing analyses remain siloed in single-omics or isolated exposures, overlooking bidirectional psychosocial feedback. By fusing genomics, proteomics, metabolomics, high-resolution environmental linkage, validated mental-health scales, and wrist-accelerometry physical-activity metrics, we will expose latent therapeutic nodes, refine risk stratification, and accelerate precision prevention. State-of-the-art causal-inference and ensemble-learning frameworks will deliver robust, clinically actionable models ready for translation
