Dr. Devlin’s research interests are in the development of adaptive and nonparametric regression methods for cancer research. With colleagues at Memorial Sloan Kettering, he recently developed a novel algorithm to refine and improve the discriminatory ability of an existing prognostic classifier by incorporating new genomic markers for censored survival data. His current research interests also include developing risk scores for high dimensional survival data when multiple markers may predict treatment response. Dr. Devlin collaborates with researchers in the Leukemia, Bone Marrow Transplant, and Myeloma Services in the design and analysis of retrospective and prospective studies.