Developments in biotechnology have extended the opportunities for therapeutics development in many disease areas. Genome-wide data can be generated for identification of molecular targets, understanding mechanism of action and identification of predictive and pharmacodynamic biomarkers. Effective utilization of these opportunities requires statisticians to develop close collaborations with basic scientists, pharmacologists, computational biologists and clinical investigators and to focus more on discovery research. In addition to providing discussion of the session presentations, I will describe experiences in expanding the Biometric Research Branch of the National Cancer Institute to incorporate computational biologists and in developing systems that facilitate therapeutics discovery in genome-wide datasets.