Learning more from microarrays: insights from modules and networks.

Abstract

Global gene expression patterns can provide comprehensive molecular portraits of biologic diversity and complex disease states, but understanding the physiologic meaning and genetic basis of the myriad gene expression changes have been a challenge. Several new analytic strategies have now been developed to improve the interpretation of microarray data. Because genes work together in groups to carry out specific functions, defining the unit of analysis by coherent changes in biologically meaningful sets of genes, termed modules, improves our understanding of the biological processes underlying the gene expression changes. The gene module approach has been used in exploratory discovery of defective oxidative phosphorylation in diabetes mellitus and also has allowed definitive hypothesis testing on a genomic scale for the relationship between wound healing and cancer and for the oncogenic mechanism of cyclin D. To understand the genetic basis of global gene expression patterns, computational modeling of regulatory networks can highlight key regulators of the gene expression changes, and many of these predictions can now be experimentally validated using global chromatin-immunoprecipitation analysis.