Statistical Research in Drug Discovery and Development

点击翻译按钮获取中文摘要

Three broad issues in drug discovery and development are studied. For target identification, use of microarrays for gene expression has become a standard practice. To test the significance of gene expression for thousands of genes, the issue of multiplicity in testing is central. A new graphical procedure is proposed. For drug discovery, the research focuses on high throughput screening (HTS). A central issue in HTS is to be able to identify as many potent drugs as possible and to achieve this within an economic timeframe. To increase the "hits", both false positive and false negative errors need to be accurately estimated and the cutoff point be chosen optimally. The major research effort focuses on the accurate estimation of false positive and false negative errors, determination of optimal cutoff points, and estimation strategy. The results will be pivotal to the planning of validation studies that are used before the screen is put into production. For drug development, it is proposed to apply modern techniques in design and analysis of experiments to pharmaceutical sciences (formulations and stability) and process R&D (chemical and biological scale-up). Two new techniques are considered: estimation of interactions in experiments with complex aliasing, and robust parameter design for process improvement. Since these techniques were developed in the context of manufacturing and hi-tech industries, new features in the pharmaceutical applications should lead to the development of new methods. The research is a jointly effort by the research group of Jeff Wu (PI) at Georgia Institute of Technology and the nonclinical biostatistics and genetics groups headed by David Stock and Kim Zerba at Bristol-Myers Squibb. Statistical tools have been widely used in drug discovery and development in the pharmaceutical industry. With the increasing competition in the industry and the explosion of disease targets, it is becoming increasingly important to have an efficient system for discovering new compounds and developing them into drugs for clinical trials and scale-up production. There has been a lot of collaborative research between academia and industries on clinical trials. Much less collaboration has been done on preclinical research like drug discovery and development. Successful implementation of this project can serve as a role model for this collaboration. It can have a major societal impact in accelerating the identification of disease target and discovery of compounds for blockbuster drugs, resulting in savings of lives and health care costs. The work should lead to new advances in theory and methodology and the research findings will be presented in professional meetings and published in trade journals. Because of the novelty of applications and the scientific relevance, the methodology and theory developed in this project will be of a broad and generic nature. They will benefit the industrial partner as well as the industry in general. Several Ph.D. students will participate in the project, splitting their time between university and industrial labs. The project will provide a new opportunity for graduate students to be exposed to cutting-edge research in drug discovery and development. It can enrich their educational experience and broaden the prospects for their careers.