D.F.T. Veiga^1*, F.F.R. Vicente^1*, M. Grivet², A. de la Fuente³ and A.T.R. Vasconcelos^1
1Laboratório Nacional de Computação Científica, Laboratório de Bioinformática, Petrópolis, RJ, Brasil
²Centro de Estudos de Telecomunicações, Pontifícia Universidade Católica do Rio de Janeiro, Rio de Janeiro, RJ, Brasil
³CRS4 Bioinformatica, Parco Tecnologico POLARIS, Pula, Italy
*These authors contributed equally to this study.
Corresponding author: A.T.R. Vasconcelos
E-mail: atrv@lncc.br

Genet. Mol. Res. 6 (4): 730-742 (2007)
Received August 03, 2007
Accepted September 25, 2007
Published October 05, 2007

ABSTRACT. Transcriptional control is an essential regulatory mechanism employed by bacteria. Much about transcriptional regulation remains to be discovered, even for the most widely studied bacterium, Escherichia coli. In the present study, we made a genome-wide low-order partial correlation analysis of E. coli microarray data with the purpose of recovering regulatory interactions from transcriptome data. As a result, we produced whole genome transcription factor regulation and co-regulation graphs using the predicted interactions, and we demonstrated how they can be used to investigate regulation and biological function. We concluded that partial correlation analysis can be employed as a method to predict putative regulatory interactions from expression data, as a complementary approach to transcription factor binding site tools and other tools designed to detect co-regulated genes.

Key words: Partial correlation, Transcriptional regulation, Co-regulation, Microarray analysis, Gene expression, Escherichia coli