Our research focuses on developing and applying computational methods to advance precision medicine, with a special aspect of incorporating evolutionary and functional information in model construction. Via close collaborations with biomedical and clinical researchers world-wide, we aim to translate bioinformatics discoveries into improvements in patient care. Our past and current research projects include:

  • Development of a novel evolution-informed modeling approach to discover biomarkers for accurate prediction of treatment outcomes. This method has been applied to leukemia, prostate cancer, breast cancer, and bladder cancer.
  • Optimization of anticancer therapy by understanding intratumor heterogeneity via subclonal reconstruction.
  • Multi-task learning to discover biomarkers for complex diseases in under-represented populations. This method has been applied to type-2 diabetes and Alzheimer’s disease.
  • Novel method for eQTL analysis with flexible LD structure and tree-guided group lasso. This method has been applied to inflammatory bowel disease.
  • Diagnosis of deleterious variants in personal exomes using evolutionary profiles. Methods have been developed for Mendelian diseases and differential drug responses.
  • Investigate evolutionary origins of human disease.
  • Assessment of biodiversity in immunoglobulin pools for autoimmune diseases.
  • Method development and analysis of metagenomics data.