ModuleBlast: identifying activated sub-networks within and across species.
Guy E Zinman, Shoshana Naiman, Dawn M O'Dee, Nishant Kumar, Gerard J Nau, Haim Y Cohen, Ziv Bar-Joseph
Author Information
Guy E Zinman: Lane Center for Computational Biology, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA 15213, USA.
Shoshana Naiman: The Mina & Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan 52900, Israel.
Dawn M O'Dee: Department of Microbiology and Molecular Genetics, School of Medicine, University of Pittsburgh, Pittsburgh, PA 15216, USA.
Nishant Kumar: Lane Center for Computational Biology, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA 15213, USA.
Gerard J Nau: Department of Microbiology and Molecular Genetics, School of Medicine, University of Pittsburgh, Pittsburgh, PA 15216, USA.
Haim Y Cohen: The Mina & Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan 52900, Israel.
Ziv Bar-Joseph: Lane Center for Computational Biology, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA 15213, USA zivbj@cs.cmu.edu.
Identifying conserved and divergent response patterns in gene networks is becoming increasingly important. A common approach is integrating expression information with gene association networks in order to find groups of connected genes that are activated or repressed. In many cases, researchers are also interested in comparisons across species (or conditions). Finding an active sub-network is a hard problem and applying it across species requires further considerations (e.g. orthology information, expression data and networks from different sources). To address these challenges we devised ModuleBlast, which uses both expression and network topology to search for highly relevant sub-networks. We have applied ModuleBlast to expression and interaction data from mouse, macaque and human to study immune response and aging. The immune response analysis identified several relevant modules, consistent with recent findings on apoptosis and NF��B activation following infection. Temporal analysis of these data revealed cascades of modules that are dynamically activated within and across species. We have experimentally validated some of the novel hypotheses resulting from the analysis of the ModuleBlast results leading to new insights into the mechanisms used by a key mammalian aging protein.
