Statistical Engineering Division Seminar

Dr. Walter Liggett
National Institute of Standards and Technology
Statistical Engineering Division
Statistical Engineering Division Seminar
Thursday May 18, 2006, 2:00-3:00 PM
Administration Building, Lecture Room C

Abstract

Using data from the Microarray Quality Control (MAQC) study, we demonstrate two data-analysis methods that shed light on the normalization of gene expression measurements and thereby on their technical variation. One is an improved method for normalization of multiple arrays with mRNA concentrations related by a parametric model. The other is a method for characterizing limitations on the effectiveness of normalization in reducing technical variation. Using our improved normalization, we consider the fit of the mixture model that relates the four materials in the study. We find that the lack of fit is significant, but small enough that its source cannot be easily identified. Applying our characterization method to measurements of the same material, we show that there is a source of variation that cannot be eliminated by normalization and therefore must be dealt with by other means. The four whole-genome, single probe, one-color microarray platforms underlie our demonstration.