Neuropsychiatric disorders are the leading cause of global disability, yet their etiology arises from a complex interplay of genetic susceptibility and modifiable environmental factors that remains incompletely understood. This project aims to address this gap by leveraging the multimodal depth of the UK Biobank to systematically identify and quantify risk factors and genetic determinants across several disorders, including depression, anxiety, bipolar disorder, schizophrenia, Parkinson’s disease, and dementia.
Our primary objectives are: 1) To identify and prioritize modifiable risk factors (e.g., pollution, sleep, cardiometabolic health, neuroimaging phenotypes) for incident disease using ensemble machine learning and time-to-event models, and to calculate population-attributable fractions to guide public health strategy; 2) To discover novel genetic susceptibility loci via exposure-stratified genome- and exome-wide association studies; 3) To model gene-environment interactions using Mendelian randomization and polygenic risk scores (PRS), elucidating how genetic risk is moderated by environmental context; 4) To develop integrated clinical risk stratification tools combining PRS, exposure risks, and neuroimaging markers for precision prevention.
The scientific rationale is that a comprehensive, data-driven integration of genomic, environmental, and phenotypic data is urgently needed to decipher differential disease vulnerability and enable targeted interventions.