Title: Integration of Functional Genomic Information: From Yeast to Worm
Speaker: Hui Ge, Harvard Medical School, Vidal Lab
Date/Time/Room: Friday, January 23, 2004 / 4:00pm / NSH 184
Host: Prof. A.-L. Barabasi

Abstract: Genomic and proteomic approaches can provide hypotheses concerning functions for the large numbers of genes predicted from genome sequences. However, information from these approaches should be considered with caution due to the artificial nature of standardized high-throughput assays. Although it is possible that biological hypotheses can be formulated by integrating functional genomic and proteomic data from various sources, it remains a question as to what extent the data can be correlated and how such integration can be achieved for different organisms. We developed strategies to relate transcriptome and interactome datasets for Saccharomyces cerevisiae and provided global evidence that genes with similar expression profiles are more likely to encode interacting proteins. To initiate studies on how interactome networks relate to multicellular functions and to extend data integration for higher organisms, we have mapped a large fraction of the Caenorhabditis elegans interactome network. More than 4000 interactions were identified by high-throughput yeast two-hybrid screens. Together with already described interactions and interologs predicted in silico, our current version of the Worm Interactome (WI5) map contains ~5500 interactions. Using this map, we generated a neighborhood of proteins essential for early embryogenesis. It was revealed on a global scale that physical interactors tend to exhibit correlated loss-of-function defects during early embryogenesis. Transcriptome and phenome information were integrated with the interactome map. Based on topological properties of the network, predictions were made for genes that are potentially involved in early embryogenesis and for gene pairs of shared functions during this process. In these predictions we recapitulated published information about early embryogenesis and provided candidate genes or gene pairs that are of highly likely to be involved in early embryogenesis. We propose that such integration of functional genomic information can be applied to other biological processes and to other organisms as well.