Statistical Methods and Applications for Microarray Studies

• Activity Number: 32
• Type: Other
• Date/Time: Sunday, August 3, 2003 : 4:00 PM to 5:50 PM
• Sponsor: ASA
• Author(s): Christiana Kendziorski*+
• Companies: University of Wisconsin, Madison

Abstract

With the sequencing of many genomes, including the human, now complete and other organisms not far behind, the era is fast approaching in which the sequences of all genes will be identified. In line with this genomic data explosion are technologies such as DNA microarrays that allow for large scale coordinate-monitoring of gene expression. Such technologies have led research efforts from analyzing individual genes to understanding the complicated and orchestrated ways in which multiple genes interact to create life. With this shift from individual to integrated analysis in molecular biology comes a shift in the related statistical questions posed and methods required. Many traditional statistical methods of data analysis do not apply to microarray data, in large part because the number of measurements of distinct genes across an array greatly exceeds that for any individual gene. As a result, new approaches to address questions particular to microarray experiments have been developed. Such questions include the normalization of measurements within and between arrays, the identification of expressed genes and estimation of true intensity, and more.