A novel approach to select differential pathways associated with hypertrophic cardiomyopathy based on gene co-expression analysis

The present study was designed to develop a novel method for identifying significant pathways associated with human hypertrophic cardiomyopathy (HCM), based on gene co-expression analysis. The microarray dataset associated with HCM (E-GEOD-36961) was obtained from the European Molecular Biology Laboratory-European Bioinformatics Institute database. Informative pathways were selected based on the Reactome pathway database and screening treatments. An empirical Bayes method was utilized to construct co-expression networks for informative pathways, and a weight value was assigned to each pathway. Differential pathways were extracted based on weight threshold, which was calculated using a random model. In order to assess whether the co-expression method was feasible, it was compared with traditional pathway enrichment analysis of differentially expressed genes, which were identified using the significance analysis of microarrays package. A total of 1,074 informative pathways were screened out for subsequent investigations and their weight values were also obtained. According to the threshold of weight value of 0.01057, 447 differential pathways, including folding of actin by chaperonin containing T-complex protein 1 (CCT)/T-complex protein 1 ring complex (TRiC), purine ribonucleoside monophosphate biosynthesis and ubiquinol biosynthesis, were obtained. Compared with traditional pathway enrichment analysis, the number of pathways obtained from the co-expression approach was increased. The results of the present study demonstrated that this method may be useful to predict marker pathways for HCM. The pathways of folding of actin by CCT/TRiC and purine ribonucleoside monophosphate biosynthesis may provide evidence of the underlying molecular mechanisms of HCM, and offer novel therapeutic directions for HCM.