Predicting gene function using DNA microarray data from multiple organisms.

摘要

The genome projects have recently established a nearly complete list of genes in several genomes, revealing that most of the genes in multicellular organisms have yet to be characterized genetically. To understand genetic mechanisms fully, one of the next challenges facing biology is to elucidate the role of every gene in the genome, especially for the large fraction of genes whose functions are currently unknown. Microarrays have created an unprecedented opportunity for investigating gene expression under a variety of different conditions on a genome-wide scale. Coexpression of genes under different conditions can provide information about gene function. However, coexpression detected in a single organism may not always imply functional relatedness.; Evolutionary conservation is a powerful criterion to identify genes that are functionally important from a set of coregulated genes. Coregulation of a pair of genes over large evolutionary distances implies that the coregulation confers a selective advantage, most likely because the genes are functionally related. This thesis describes a method for identifying evidence of such conserved gene coexpression from DNA microarray data.; Applying the method to the compendiums of DNA microarray data from human, fly, worm, and yeast identified a network containing 3416 genes linked by 22,163 conserved interactions. Many of these genes in the resulting network are uncharacterized and therefore the results provide an opportunity for predicting their functions for the first time. Several areas of highly interconnected genes were found in the network, revealing an overall structure present among the coexpression links. Computational evaluations and experimental verification demonstrate that the multiple species approach outperforms those based on expression data from only a single organism. Combining microarray data across multiple organisms can therefore improve gene function prediction and provides clues about the way conserved genetic modules evolve.