Machine Learning Algorithm-Based Discovery of Potential Regulators of Immune-Related Dilated Cardiomyopathy

Dilated cardiomyopathy (DCM) is considered a severe non-ischemic myocardial disease, and there is currently no effective method for the early detection of DCM. Therefore, we aimed to use machine learning algorithms to discover more accurate factors to guide clinical drug development and precision medicine diagnosis.

Two datasets containing patients with DCM and healthy controls were downloaded from the Gene Expression Omnibus database. After data preprocessing, differentially expressed genes (DEGs) between the DCM patients and normal samples were identified using the limma package. In addition, to screen for DEGs closely associated with immune inflammation, we collected immune-related genes and defined overlapping genes as differential immune genes (Immune-DEGs). Protein-protein interaction (PPI) network construction and functional enrichment analysis were then functionally validated for the differential immune genes. Subsequently, we further screened the immune-DEGs using the least absolute shrinkage and selection operator (LASSO) technique and support vector machine algorithm (SVM), resulting in the screening of five potential modulators closely associated with DCM. Finally, the diagnostic efficacy of the modifiers was assessed using subject operating characteristic curves based on independent external data, and the intrinsic pathological mechanisms of different differential immune genes were explored by immune infiltration analysis.

A consensus of 184 differential immune genes were identified, and the functional enrichment results of their PPI network modules suggested that inflammation, immune disorders, and viral infections play an essential role in the pathogenesis of DCM. Five signature genes were then further screened using LASSO and SVM algorithms: KLRC4, CCL4, IGHV3-33, ITGAL, and inducible T-cell kinase.

This study constructed a gene set of potential DCM regulators with five immune-related genes, which could provide a new strategy for the diagnosis and treatment of DCM.