Our laboratory uses genomics approaches to investigate the pathogenesis and biology of human cancer, and to identify strategies for improved cancer diagnosis, prognostication and patient management. In addition to expression-profiling studies, we have pioneered array-based comparative genomic hybridization (array CGH) methods to map DNA copy number alterations genome-wide at high-resolution in tumor genomes, and apply biocomputational approaches to integrate array CGH and expression-profiling data to discover additional insights into cancer pathogenesis. Active areas of investigation include: (1) Defining gene-expression signatures for improved clinical outcome-prediction and risk-assessment in acute myeloid leukemia; (2) Characterizing newly discovered gene-expression subtypes of prostate cancer to identify pathogenetic alterations and biomarkers for risk-assessment; (3) Mapping recurrent DNA amplifications and deletions in common epithelial tumors (breast, pancreatic and lung cancers) to discover new cancer genes; (4) Investigating mechanisms underlying genomic instability and the shaping of tumor genomes.