I am a bioinformatician with training in mathematics, machine learning, genetics and digital medicine. Our lab is mainly focused on the development and application of computational tools to study complex diseases. The ultimate goal is to develop new computational strategies for the disease prevention and intervention.
- Identification of genetic causes of complex diseases. We have been involved in multiple large-scale genetic consortiums, such as the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium, Trans-Omics for Precision Medicine (TOPMed) program, and Alzheimer's Disease Sequencing Project (ADSP). These studies have identified hundreds of genetic loci associated with atrial fibrillation, heart failure, hypertension and Alzheimer’s disease.
- Integration of multi-omics data to understand disease molecular mechanisms. Complex diseases are usually caused by the interplay of genetic and environmental factors. We have identified numerous molecular signatures from gene expression, protein expression, and DNA methylation that are related to aging and cardiovascular disease. We are also developing computational methods to integrate different molecular signatures and build gene interaction networks to study potential disease regulation networks.
- Development of machine learning models for the early disease diagnosis. We have built multiple machine learning models to predict dementia risk from midlife risk factors and neuropsychological tests. In combination with neuroimaging and blood-based measures, we are also developing multimodal machine learning methods to identify new biomarkers that are predictive of future cognitive impairment.
- Exploration of digital and wearable devices for health monitoring. We have deployed thousands of wearable devices and mobile apps to monitor cardiovascular health and cognitive health. We are integrating active engagement with passive engagement technologies from the habitual environment to make sustained monitoring feasible. Novel analytic strategies are also being developed to analyze big unstructured data to identify potential digital biomarkers that are predictive of future health outcomes.