Coronary artery disease associated specific modules and feature genes revealed by integrative methods of WGCNA, MetaDE and machine learning.

PURPOSE

Coronary artery disease (CAD) is one of the most common causes of morbidity and mortality globally. This work aimed to investigate the specific modules and feature genes associated with CAD.

METHODS

Three microarray datasets were downloaded from the Gene Expression Omnibus database, which included CAD and healthy samples. WGCNA was applied to identify highly preserved modules across the three datasets. MetaDE method was used to select differentially expressed genes (DEGs) with significant consistency. Protein-protein interaction (PPI) network was constructed using the overlapping genes amongst the DEGs with significant consistency and in the preserved modules. Moreover, a combined machine learning of support vector machine and recursive feature elimination was used to further investigate the feature genes and pathways.

RESULTS

Nine highly preserved modules were detected in the WGCNA network, and 961 DEGs with significant consistency across the three datasets were selected using the metaDE method. A PPI network was constructed with the 158 overlapping genes. Ten genes were found to be involved in these KEGG pathways directly, including genes CD22, CD79B, CD81, CR1, IKBKE, MAP3K3, MAPK14, MMP9, NCF4, and SPP1.

CONCLUSIONS

The present work might provide novel insight into the underlying molecular mechanism of CAD.