Investigating Dysregulated Pathways in Dilated Cardiomyopathy from Pathway Interaction Network

Abstract

The aim of this study was to identify dysregulated pathways for the diagnosis and treatment of dilated cardiomyopathy (DCM) using pathway interaction network analysis. Methods: Transcriptome data of DCM, protein-protein interaction (PPI) data, and pathway data were recruited and preprocessed. Then, the pathway interaction network was constructed based on the gene expression analysis and gene co-expression analysis. Meanwhile, pathway activity analysis was performed, and the pathway with the greatest activity change was defined as the seed pathway. Staring from the seed pathway, the dysregulated pathways that could serve as diagnostic biomarker was extracted from the pathway interaction network using support vector machines. Results: Combining gene expression and co-expression data, we constructed the pathway interaction network, covering 4175 pathway interactions. Via pathway activity analysis, cap-dependent translation initiation with the greatest activity change was defined as the seed pathway. Staring from cap-dependent translation initiation, a total of 21 dysregulated pathways were obtained, which could discriminate DCM samples from controls with the area under the curve value of 0.95. Conclusion: A pathway interaction network was implemented to identify dysrgulated pathways that can best discriminate DCM samples from controls. We identified a total of 21 dysregulated pathways in DCM, which can serve as diagnostic biomarkers for DCM.