We are developing methods to integrate genome sequence annotation data with microarray expression data. This effort has two emphases: (i) comparison of clusters of expressed genes with function, genomic location and patterns of evolutionary conservation of proteins from all fully sequenced genomes; and (ii) automatic annotation of cDNAs corresponding to microarray positions. Our goal is to generate explanations of gene clusters built from gene expression data. To construct explanations, we draw on databases of metabolic pathway information, genome annotation and genomic organization (prokaryotic operons). A preliminary version of our system is implemented as a cDNA annotation module of the MAGPIE genome annotation system. This talk will present the use of the cDNA annotation system for Xenopus, mouse hypothalamus and Phretameoba cDNA sequences.