Research Question and Objectives
Chronic diseases and cancers (e.g., cardiovascular diseases, kidney disease, type II diabetes, lung and colorectal cancer) are leading global health burdens, exacerbated by aging populations and environmental pollution. While genetic and lifestyle factors are partially understood, interactions between environmental exposures (air pollution, heavy metals), metabolic dysregulation, and molecular mechanisms remain unclear. Leveraging UK Biobank’s multi-omics data , environmental records, and lifestyle metrics, this study aims to:
1.Identify novel biomarkers and environmental risk factors via integrated omics-environment analysis;
2.Decipher causal pathways linking exposures to molecular changes (e.g., inflammation, oxidative stress, epigenetics) and disease;
3.Develop machine learning models for early risk stratification and precision intervention.
Scientific Rationale
Circulating proteins (e.g., IL-6, CRP) and metabolites (e.g., branched-chain amino acids, lipid derivatives) mediate chronic disease and cancer. PM2.5 impairs mitochondrial complex I, causing oxidative damage that accelerates atherosclerosis and insulin resistance. Arsenic promotes tumorigenesis via p53 methylation and HIF-1!/VEGF-driven angiogenesis. UK Biobank provides high-resolution exposure data (land-use regression for PM2.5/NO!, British Geological Survey arsenic levels) and objective physical activity measures from accelerometers. Mendelian randomization will use GWAS-identified PM2.5-associated SNPs as instruments to infer causality, minimizing confounding. Multi-omics mediation models will test the pathway: exposure ! metabolic dysregulation ! epigenetic change ! disease. To enhance clinical utility, random forest and XGBoost models with SHAP interpretation will identify key predictors , overcoming limitations of single-omics approaches. Longitudinal follow-up enables robust risk prediction and mechanistic inference.