Predicting gene regulation by sigma factors in Bacillus subtilis from genome-wide data

Motivation: Sigma factors regulate the expression of genes in Bacillus subtilis at the transcriptional level. First we assess the ability of currently available gene regulatory network models to accurately infer gene regulation by sigma factors from gene expression data. Secondly, we consider improving the prediction accuracy by combining gene expression data with sequence information. Finally, we apply the resulting joint predictor to discover currently unknown gene regulations by sigma factors in Bacillus subtilis. Methods: We determine the accuracy of sigma factor prediction from gene expression data using a foldchange analysis, Bayesian networks, dynamic models, and supervised learning based on coregulation. We show that the recently proposed method of combining a coregulation-based prediction with sequence information by summing the log-likelihood scores (Segal et al., 2003), at least in our case, effectively ignores sequence information. We propose to use logistic regression to ...

• Type: Conference
• Year: 2004
• Where: ISMB
• Authors: Michiel J. L. de Hoon, Yuko Makita, Seiya Imoto, Kazuo Kobayashi, Naotake Ogasawara, Kenta Nakai, Satoru Miyano
• Keywords: Computational Biology, Gene Expression Data, ISMB 2004, Sequence Information, Sigma Factors