We develop computational methods for studying renal biopsies of patients with proteinuria to precisely identify early glomerular structural changes, and study whether the resulting computationally derived features are associated with disease progression.  Primary goal is to identify patients at risk of renal failure.  As part of this effort, we focus on diabetic nephropathy as disease model, and we also study renal transplant cases. 

We also develop computational methods to study cell structure and function across cell cultures, tissues, and species using image based features.  Goal is to translate findings from cell culture systems to animal model systems, as well as from animal model systems to human systems.  As part of this effort, we focus on studying renal cell biology of murine and human systems.