We develop methods to overcoming time-dependent confounding affected by previous treatments in observational studies. This includes novel marginal structural models, their associated theoretical properties, and tests of their performance using Monte Carlo simulation studies. We also research methods for combining longitudinal matching with missing data methods, variance estimation, and mediation analyses for rare disease studies with limited sample size. New models are applied to real data from observational cohort studies and electronic health record databases to answer important clinical questions for children and adults with kidney diseases. These projects are supported by an NIH/NIDDK-funded P50 grant and were previously supported by a NephCure pilot grant.
We collaborate with nephropathologists, nephrologists, engineers, and molecular biologists to extract information from digital kidney biopsy images and identify important diagnostic, mechanistic, and prognostic imaging features. We develop novel high-dimensional statistical methods to analyze these data and use machine learning, computational image analysis, and bioinformatics tools to extract and integrate (bulk, single-cell, and spatial) transcriptomic, imaging, and clinical data. These projects are supported by multiple NIH/NIDDK-funded R01 grants.
We collaborate with investigators in the Nephrotic Syndrome Study Network (NEPTUNE) and Cure Glomerulonephropathy Network (CureGN) to conduct clinical research in glomerular disease, including projects related to biomarker discovery, nocturnal blood pressure dysregulation, vaccine effectiveness, pregnancy, and skeletal outcomes. Our research team leads studies and develops new methods for analyses of time-varying biomarkers and kidney function outcomes. These projects are supported by an NIH/NIDDK-funded R01 grant, an NIH/NHLBI-funded R01 grant, and the NIH/NIDDK-funded NEPTUNE U54 grant.
We collaborate with pediatric nephrologists at the Children's Hospital of Philadelphia to conduct clinical research in pediatric kidney diseases, including acute kidney injury, chronic kidney disease, kidney transplant, and thrombotic microangiopathy. We lead research on kidney disease progression as part of the Preserving Kidney Function in Children with Chronic Kidney Disease (PRESERVE) study that leverages the PCORnet electronic health records database. These projects are supported by an NIH/NHLBI-funded R01 grant and the Penn-CHOP Kidney Innovation Center.
University of Pennsylvania
University of Pennsylvania
University of North Carolina Chapel Hill
Duke University
University of Michigan
Emory University and Georgia Tech
Children's Hospital of Philadelphia and University of Pennsylvania
University of California Davis