Finding Common Modules in a Time-Varying Network with Application to the Drosophila Melanogaster Gene Regulation Network

ABSTRACT

Finding functional modules in gene regulation networks is an important task in systems biology. Many methods have been proposed for finding communities in static networks; however, the application of such methods is limited due to the dynamic nature of gene regulation networks. In this article, we first propose a statistical framework for detecting common modules in the Drosophila melanogaster time-varying gene regulation network. We then develop both a significance test and a robustness test for the identified modular structure. We apply an enrichment analysis to our community findings, which reveals interesting results. Moreover, we investigate the consistency property of our proposed method under a time-varying stochastic block model framework with a temporal correlation structure. Although we focus on gene regulation networks in our work, our method is general and can be applied to other time-varying networks. Supplementary materials for this article are available online.

KEYWORDS:

• Community structure
• Consistency
• Gene regulation network
• Hypothesis testing
• Markov chain Monte Carlo
• Robustness

Supplementary Materials

The supplementary materials contain proof of Theorem 1, details of the genes contained in the identified communities, further information on the gene ontology enrichment analysis and community findings when applying Method 2 to the Drosophila melanogaster gene regulation network.

Acknowledgments

The authors are grateful to Professor Eric P. Xing and Professor Le Song for providing the gene regulation network data. The authors thank Professor Bingyun Sun for her valuable input on our enrichment analysis and Michael Grosskopf, Christina Nieuwoudt, and Yunlong Nie for their proofreading. The authors are very grateful for the constructive comments from the editor, associate editor, and two anonymous reviewers. These comments have helped greatly in improving this work.

Funding

This research is partially supported by a discovery grant from the Natural Sciences and Engineering Research Council of Canada (NSERC) to the second author.