1 INTRODUCTION In microarray studies, to identify genes that are associated with an outcome of interest, a large number of gene expressions (potentially tens of thousands) are measured from typically a much smaller number of subjects (often in the tens to hundreds). These gene expression measurements tend to be noisy, with measurement errors coming from many sources such as sample preparation, labeling, and hybridization; for example, see Rocke and Durbin (2001) and Zakharkin et al. (2005).
