Research

Integrative modeling of molecular regulatory networks. 

We develop computational methods that integrate multiple layers of molecular information to reconstruct gene regulatory networks underlying complex diseases. By combining complementary genomic and transcriptomic data with clinical and genetic evidence, these approaches identify biologically meaningful regulatory mechanisms and reveal disease-specific pathways across heterogeneous patient populations. 

Network medicine and graph-based AI for therapeutic discovery. 

We develop graph representation learning and network medicine approaches that integrate molecular, pharmacologic, and clinical information to predict therapeutic response and identify opportunities for drug repurposing. These methods model patients, drugs, and biological pathways within unified network frameworks, enabling individualized treatment prediction while bridging the gap between experimental models and clinical populations. 

Statistical genetics and AI for population-scale disease genomics. 

We develop computational and statistical methods for analyzing large-scale biobank and EHR-linked genomic data to characterize the genetic architecture of complex diseases across diverse populations. This research focuses on scalable association testing, polygenic risk modeling, and subgroup-aware inference, with an emphasis on translating genomic discoveries into clinically meaningful insights.