Statistical Proteomics in Chemoprevention Medical Research
by
Mark Carpenter
Department of Discrete and Statistical Sciences, Auburn University, Auburn, Alabama
Coauthors: Craig Rowell, Coral Lamartiniere (Department of Pharmacology and Toxicology, University of Alabama at Birmingham, Birmingham, Alabama 35294)

Modern oroteomics is the science of making quantitative comparisons of proteomes under different conditions, e.g., normal versus cancerous tissues or treated versus untreated tissues, to understand biological processes and/or the nature of a disease. Because the human proteome is encoded by the human genome it is broader in its complexities. For example, while the human genome comprises approximately 30,000 genes, it is estimated that the human proteome may comprise 500,000 proteins. The development of statistical, computational and bioinformatical methods is key to unlocking the mysteries of the human proteome. Not surprisingly, the emerging proteomic technologies designed to study various proteomes produce immense challenges to the statistician. Because each tissue or blood sample can express many thousands of proteins, the data is highly-dimensional with any given experiment generating potentially millions of data elements, even for small sample sizes. In this paper, we proffer statistical and mathematical methods for the identification, quantification, and differentiation of proteins based on 2D-gel technology and mass spectrometry.

Date received: September 19, 2003