This article is part of the supplement: Proceedings of the 21st International Conference on Genome Informatics (GIW2010)
A novel parametric approach to mine gene regulatory relationship from microarray datasets
- Equal contributors
1 State Key Laboratory of Proteomics, Beijing Proteome Research Center, Beijing Institute of Radiation Medicine, Beijing 102206, China
2 Department of Chemistry and Biology, College of Science, National University of Defense Technology, Changsha 410073, China
BMC Bioinformatics 2010, 11(Suppl 11):S15 doi:10.1186/1471-2105-11-S11-S15Published: 14 December 2010
Microarray has been widely used to measure the gene expression level on the genome scale in the current decade. Many algorithms have been developed to reconstruct gene regulatory networks based on microarray data. Unfortunately, most of these models and algorithms focus on global properties of the expression of genes in regulatory networks. And few of them are able to offer intuitive parameters. We wonder whether some simple but basic characteristics of microarray datasets can be found to identify the potential gene regulatory relationship.
Based on expression correlation, expression level variation and vectors derived from microarray expression levels, we first introduced several novel parameters to measure the characters of regulating gene pairs. Subsequently, we used the naïve Bayesian network to integrate these features as well as the functional co-annotation between transcription factors and their target genes. Then, based on the character of time-delay from the expression profile, we were able to predict the existence and direction of the regulatory relationship respectively.
Several novel parameters have been proposed and integrated to identify the regulatory relationship. This new model is proved to be of higher efficacy than that of individual features. It is believed that our parametric approach can serve as a fast approach for regulatory relationship mining.