ABSTRACT
Pooling data across multiple microarray studies should provide increased power to detect new relationships between gene expression levels and outcomes of interest. In the chapter of Goldstein et al. (Chapter 9), the data from two different microarray studies are derived from the same type, making pooling relatively straightforward. In general, though, pooling is complicated by the fact that gene expression measurements from different microarray platforms are not directly comparable. In this chapter, we discuss two methods for combining information across different versions of Affymetrix oligonucleotide arrays (Lockhart et al., 1996; see also Section 1.4). Each involves a new approach for combining probes on the array into probe sets.
