Limits...
Controlling the response: predictive modeling of a highly central, pathogen-targeted core response module in macrophage activation.

McDermott JE, Archuleta M, Thrall BD, Adkins JN, Waters KM - PLoS ONE (2011)

Bottom Line: This module occupies a highly central position in the inferred network and is also enriched in genes preferentially targeted by pathogens.The module includes cytokines, interferon induced genes such as Ifit1 and 2, effectors of inflammation, Cox1 and Oas1 and Oasl2, and transcription factors including AP1, Egr1 and 2 and Mafb.Predictive modeling using a reverse-engineering approach reveals dynamic differences between the responses to each stimulus and predicts the regulatory influences directing this module.

View Article: PubMed Central - PubMed

Affiliation: Computational Biology and Bioinformatics, Pacific Northwest National Laboratory, Richland, Washington, United States of America. Jason.McDermott@pnl.gov

ABSTRACT
We have investigated macrophage activation using computational analyses of a compendium of transcriptomic data covering responses to agonists of the TLR pathway, Salmonella infection, and manufactured amorphous silica nanoparticle exposure. We inferred regulatory relationship networks using this compendium and discovered that genes with high betweenness centrality, so-called bottlenecks, code for proteins targeted by pathogens. Furthermore, combining a novel set of bioinformatics tools, topological analysis with analysis of differentially expressed genes under the different stimuli, we identified a conserved core response module that is differentially expressed in response to all studied conditions. This module occupies a highly central position in the inferred network and is also enriched in genes preferentially targeted by pathogens. The module includes cytokines, interferon induced genes such as Ifit1 and 2, effectors of inflammation, Cox1 and Oas1 and Oasl2, and transcription factors including AP1, Egr1 and 2 and Mafb. Predictive modeling using a reverse-engineering approach reveals dynamic differences between the responses to each stimulus and predicts the regulatory influences directing this module. We speculate that this module may be an early checkpoint for progression to apoptosis and/or inflammation during macrophage activation.

Show MeSH

Related in: MedlinePlus

Overview of computational approaches.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3038849&req=5

pone-0014673-g001: Overview of computational approaches.

Mentions: Our overall goal in this study was to characterize the similarities between macrophage responses to multiple stimuli, including an intracellular bacteria (S. Typhimurium), and inert particles, and to identify important regulatory influences in macrophage activation. To accomplish this we used several different computational approaches (Figure 1). First we inferred regulatory association networks using the Context Likelihood of Relatedness (CLR) method [19]. CLR establishes relationships (edges in a graph) between two genes when the expression of one gene has significant mutual information (i.e. highly similar or dissimilar) with the expression of another gene. The resulting networks summarize the functional dynamics of the system, for the conditions considered. We used this coexpression network to predict important regulatory influences using topological analysis. We then compared the responses of macrophages to a number of important stimuli; TLR agonists, bacterial infection, and inert nanoparticle exposure. We used this analysis to identify a set of genes that was differentially regulated under all conditions examined. To understand the regulation of this core response module we used a multivariate regression method to develop a model of the regulatory influences of the module. This model was validated by assessing its ability to predict gene expression under novel conditions. We finally discuss the results of this analysis in terms of biological insight offered into macrophage activation.


Controlling the response: predictive modeling of a highly central, pathogen-targeted core response module in macrophage activation.

McDermott JE, Archuleta M, Thrall BD, Adkins JN, Waters KM - PLoS ONE (2011)

Overview of computational approaches.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3038849&req=5

pone-0014673-g001: Overview of computational approaches.
Mentions: Our overall goal in this study was to characterize the similarities between macrophage responses to multiple stimuli, including an intracellular bacteria (S. Typhimurium), and inert particles, and to identify important regulatory influences in macrophage activation. To accomplish this we used several different computational approaches (Figure 1). First we inferred regulatory association networks using the Context Likelihood of Relatedness (CLR) method [19]. CLR establishes relationships (edges in a graph) between two genes when the expression of one gene has significant mutual information (i.e. highly similar or dissimilar) with the expression of another gene. The resulting networks summarize the functional dynamics of the system, for the conditions considered. We used this coexpression network to predict important regulatory influences using topological analysis. We then compared the responses of macrophages to a number of important stimuli; TLR agonists, bacterial infection, and inert nanoparticle exposure. We used this analysis to identify a set of genes that was differentially regulated under all conditions examined. To understand the regulation of this core response module we used a multivariate regression method to develop a model of the regulatory influences of the module. This model was validated by assessing its ability to predict gene expression under novel conditions. We finally discuss the results of this analysis in terms of biological insight offered into macrophage activation.

Bottom Line: This module occupies a highly central position in the inferred network and is also enriched in genes preferentially targeted by pathogens.The module includes cytokines, interferon induced genes such as Ifit1 and 2, effectors of inflammation, Cox1 and Oas1 and Oasl2, and transcription factors including AP1, Egr1 and 2 and Mafb.Predictive modeling using a reverse-engineering approach reveals dynamic differences between the responses to each stimulus and predicts the regulatory influences directing this module.

View Article: PubMed Central - PubMed

Affiliation: Computational Biology and Bioinformatics, Pacific Northwest National Laboratory, Richland, Washington, United States of America. Jason.McDermott@pnl.gov

ABSTRACT
We have investigated macrophage activation using computational analyses of a compendium of transcriptomic data covering responses to agonists of the TLR pathway, Salmonella infection, and manufactured amorphous silica nanoparticle exposure. We inferred regulatory relationship networks using this compendium and discovered that genes with high betweenness centrality, so-called bottlenecks, code for proteins targeted by pathogens. Furthermore, combining a novel set of bioinformatics tools, topological analysis with analysis of differentially expressed genes under the different stimuli, we identified a conserved core response module that is differentially expressed in response to all studied conditions. This module occupies a highly central position in the inferred network and is also enriched in genes preferentially targeted by pathogens. The module includes cytokines, interferon induced genes such as Ifit1 and 2, effectors of inflammation, Cox1 and Oas1 and Oasl2, and transcription factors including AP1, Egr1 and 2 and Mafb. Predictive modeling using a reverse-engineering approach reveals dynamic differences between the responses to each stimulus and predicts the regulatory influences directing this module. We speculate that this module may be an early checkpoint for progression to apoptosis and/or inflammation during macrophage activation.

Show MeSH
Related in: MedlinePlus