My general interest focuses on statistical and computational approaches to extract knowledge from various experimental data to understand diseases and facilitate diagnosis and intervention. Currently, we are mainly working on statistical methods to (1) improve the accuracy of the protein identification/quantification and reliability of the FDR estimate in tandem mass spectrometry and (2) integrate gene expression, transcription-factor-DNA binding, epigenetics and other information to infer the mechanism of transcriptional regulation. A major application of above efforts is to understand the heterogeneity of breast cancer etiology for more accurate and informative diagnosis and effective treatment.

We are also interested in developing statistical methods for the analysis or design of observational/interventional studies that are subjected to missing values. To date, this effort has been mostly focused on analysis of longitudinal cohort studies of dementia including Alzheimer’s disease in order to identify risk factors.