Research focuses on probabilistic modeling and machine learning for genomic medicine: calibrating high-throughput functional assays, prioritizing genes/variants, and estimating class priors using PU learning and network methods to support rare disease diagnosis.