Limits...
Identification of differentially expressed subnetworks based on multivariate ANOVA.

Hwang T, Park T - BMC Bioinformatics (2009)

Bottom Line: Our approach was successfully applied to human microarray datasets.Each identified subnetwork was annotated with the Gene Ontology (GO) term, resulting in the phenotype-related functional pathway or complex.We also compared these results with those of other scoring methods such as t statistic- and mutual information-based scoring methods.

View Article: PubMed Central - HTML - PubMed

Affiliation: Interdisciplinary Program of Bioinformatics, Seoul National University, Seoul, Republic of Korea. hwangty@snu.ac.kr

ABSTRACT

Background: Since high-throughput protein-protein interaction (PPI) data has recently become available for humans, there has been a growing interest in combining PPI data with other genome-wide data. In particular, the identification of phenotype-related PPI subnetworks using gene expression data has been of great concern. Successful integration for the identification of significant subnetworks requires the use of a search algorithm with a proper scoring method. Here we propose a multivariate analysis of variance (MANOVA)-based scoring method with a greedy search for identifying differentially expressed PPI subnetworks.

Results: Given the MANOVA-based scoring method, we performed a greedy search to identify the subnetworks with the maximum scores in the PPI network. Our approach was successfully applied to human microarray datasets. Each identified subnetwork was annotated with the Gene Ontology (GO) term, resulting in the phenotype-related functional pathway or complex. We also compared these results with those of other scoring methods such as t statistic- and mutual information-based scoring methods. The MANOVA-based method produced subnetworks with a larger number of proteins than the other methods. Furthermore, the subnetworks identified by the MANOVA-based method tended to consist of highly correlated proteins.

Conclusion: This article proposes a MANOVA-based scoring method to combine PPI data with expression data using a greedy search. This method is recommended for the highly sensitive detection of large subnetworks.

Show MeSH

Related in: MedlinePlus

The structures of the PPI network for the simulation study. (a) An overall PPI network for the simulation study. (b) Five types of target subnetworks. The red nodes represent the seeds. The five types of network structures in (b) are assumed to be target networks for the PPI network (a).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2696448&req=5

Figure 3: The structures of the PPI network for the simulation study. (a) An overall PPI network for the simulation study. (b) Five types of target subnetworks. The red nodes represent the seeds. The five types of network structures in (b) are assumed to be target networks for the PPI network (a).

Mentions: We performed a series of simulation studies to evaluate the three scoring methods. Specifically, we focused on the ability of each method to identify previously assumed target subnetworks. The simulation study was performed for the PPI network in Figure 3a, which has proteins within a depth of two from a given seed protein. For this PPI network, five types of network structures were assumed to be true target networks (Figure 3b). Expression values with correlation coefficient ρ were assigned to the nodes of the target networks, while expression values with correlation coefficient ρ' were assigned to the other remaining nodes. The expression values with correlation coefficient ρ and ρ' were generated by the following procedure.


Identification of differentially expressed subnetworks based on multivariate ANOVA.

Hwang T, Park T - BMC Bioinformatics (2009)

The structures of the PPI network for the simulation study. (a) An overall PPI network for the simulation study. (b) Five types of target subnetworks. The red nodes represent the seeds. The five types of network structures in (b) are assumed to be target networks for the PPI network (a).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2696448&req=5

Figure 3: The structures of the PPI network for the simulation study. (a) An overall PPI network for the simulation study. (b) Five types of target subnetworks. The red nodes represent the seeds. The five types of network structures in (b) are assumed to be target networks for the PPI network (a).
Mentions: We performed a series of simulation studies to evaluate the three scoring methods. Specifically, we focused on the ability of each method to identify previously assumed target subnetworks. The simulation study was performed for the PPI network in Figure 3a, which has proteins within a depth of two from a given seed protein. For this PPI network, five types of network structures were assumed to be true target networks (Figure 3b). Expression values with correlation coefficient ρ were assigned to the nodes of the target networks, while expression values with correlation coefficient ρ' were assigned to the other remaining nodes. The expression values with correlation coefficient ρ and ρ' were generated by the following procedure.

Bottom Line: Our approach was successfully applied to human microarray datasets.Each identified subnetwork was annotated with the Gene Ontology (GO) term, resulting in the phenotype-related functional pathway or complex.We also compared these results with those of other scoring methods such as t statistic- and mutual information-based scoring methods.

View Article: PubMed Central - HTML - PubMed

Affiliation: Interdisciplinary Program of Bioinformatics, Seoul National University, Seoul, Republic of Korea. hwangty@snu.ac.kr

ABSTRACT

Background: Since high-throughput protein-protein interaction (PPI) data has recently become available for humans, there has been a growing interest in combining PPI data with other genome-wide data. In particular, the identification of phenotype-related PPI subnetworks using gene expression data has been of great concern. Successful integration for the identification of significant subnetworks requires the use of a search algorithm with a proper scoring method. Here we propose a multivariate analysis of variance (MANOVA)-based scoring method with a greedy search for identifying differentially expressed PPI subnetworks.

Results: Given the MANOVA-based scoring method, we performed a greedy search to identify the subnetworks with the maximum scores in the PPI network. Our approach was successfully applied to human microarray datasets. Each identified subnetwork was annotated with the Gene Ontology (GO) term, resulting in the phenotype-related functional pathway or complex. We also compared these results with those of other scoring methods such as t statistic- and mutual information-based scoring methods. The MANOVA-based method produced subnetworks with a larger number of proteins than the other methods. Furthermore, the subnetworks identified by the MANOVA-based method tended to consist of highly correlated proteins.

Conclusion: This article proposes a MANOVA-based scoring method to combine PPI data with expression data using a greedy search. This method is recommended for the highly sensitive detection of large subnetworks.

Show MeSH
Related in: MedlinePlus