Bayesian network expansion identifies new ROS and biofilm regulators.

Area of Science:

• Systems Biology
• Computational Biology
• Genomics

Background:

• Gene expression pathways are incompletely understood in most organisms.
• Microarray data offers potential for reconstructing regulatory networks and identifying novel interactions.

Purpose of the Study:

• To reconstruct and expand the reactive oxygen species (ROS) pathway in Escherichia coli using gene expression data.
• To develop and validate a computational method for discovering unknown pathway components and interactions.

Main Methods:

• Construction of Bayesian network models from a compendium of Escherichia coli microarray data.
• Generation of a consensus network and expansion using the novel 'BN+1' algorithm.
• Experimental validation of predicted genes and interactions using reporter strains and gene knockout experiments.

Main Results:

• A Bayesian network model of the ROS pathway was generated, revealing discrepancies with known pathways.
• The BN+1 algorithm predicted stress-related genes, including uspE and gadX, and their potential interactions.
• Experimental validation confirmed uspE's role in the ROS pathway and a feedback loop between uspE and gadX, both regulating biofilm formation.

Conclusions:

• The BN+1 expansion method effectively identifies previously unknown genes within biological pathways.
• This computational approach can uncover significant biological roles for newly identified genes and interactions.
• The BN+1 method is a generalizable tool for pathway characterization and expansion in diverse biological systems.