Research Areas

Genetic architecture of neuropsychiatric disorders

Neuropsychiatric disorders have a clear genetic basis and pose an increasing societal burden, with considerable disparities in research and healthcare across populations. While genome-wide association studies (GWAS) have revealed associated genetic loci, much remains to be untangled underlying the heterogeneous symptoms and complex comorbidities of these disorders. Our lab studies the genetic determinants of psychiatric (e.g., bipolar disorder) and neurological disorders (e.g., epilepsy) by analyzing both common and rare variations and their interplay from DNA sequence data. We functionally characterize these findings, examine the polygenicity and pleiotropy patterns, and conduct cross-ancestry comparisons to identify distinct and shared genetic architectures.

Genomics of disease trajectories throughout the life course

While GWAS have predominantly studied disease susceptibility using case-control designs, the trajectory of diseases encompasses a spectrum of medical conditions and outcomes where genetics may play a role in informing precision prevention and care. We leverage large-scale biobanks and national insurance claims data to understand how genetic variations influence various disease endpoints through a time-to-event approach, such as age of onset, severity, and treatment response. We also address potential biases (e.g., colliders) in analyzing disease prognosis and progression phenotypes to derive valid causal genetic effect estimates.

Applications and prediction modeling with polygenic scores

Polygenic scores (PGS) measure individual genetic liability to a trait and have become an increasingly powerful genomic predictor that can distinguish between high and low disease risk individuals. Our interest in PGS applications is multi-fold: to investigate and improve the transferability of PGS prediction for complex diseases across ancestries; to develop enhanced genomic and multi-omic prediction models incorporating non-linear and epistatic effects using artificial intelligence; to assess disease burden and cumulative risk attributable to PGS in prospective population and clinical settings; and to dissect the contributions of genetic vs. non-genetic risk factors to disease through an integrative approach.