Identification of Exercise Signature Genes Potentially Associated with Major Depressive Disorder by Combining Bioinformatics, Machine Learning, and Mendelian Randomization Analysis

Abstract

Abstract Purpose: Aberrant alterations in exercise-related genes can significantly increase the vulnerability to major depressive disorder (MDD). This study aimed to systematically screen and identify exercise genes associated with the onset of MDD. Method: The Gene Expression Omnibus and GeneCards databases were jointly utilized to screen for exercise differentially expressed genes (DEGs) in the depression dataset GSE38206. Functional enrichment and immunological analyses of DEGs were performed subsequently. Following this, the identification of signature genes was carried out using the Least Absolute Shrinkage and Selection Operator and Random Forest analysis methods. These signature genes were then subjected to validation for diagnostic efficacy, genome enrichment analysis, and immunological characterization. Finally, a two-sample Mendelian randomization analysis was conducted to confirm the causal relationship between physical exercise and MDD. Results: A total of 256 differentially DEGs were identified through variance analysis. GeneCards was used to screen 962 exercise-related genes, resulting in the identification of 23 exercise DEGs through intersection. Exercise DEGs are enriched in signaling pathways such as advanced glycosylation end-product-receptor, tumor necrosis factor, and nuclear factor-κB, and are involved in physiological processes like response to lipopolysaccharide-mediated signaling pathways, negative regulation of exogenous apoptosis, and regulation of leukocyte immunity. Machine learning identified two signature genes, SNORD15A and RMRP, with areas under the receiver operating characteristic curve of 0.994 and 0.998, respectively. Enrichment analysis results indicate that RMRP regulates signaling pathways such as endocytosis, pyrimidine metabolism, ribosomal and protein hydrolysis, and is responsible for macroautophagy, phospholipid biosynthesis, immunomodulation, and nucleic acid metabolism. SNORD15A exhibits enrichment in signaling pathways related to extracellular matrix and receptor interactions, glycerophospholipid metabolism, as well as ribosomal and toll-like receptors. It is intricately involved in regulatory mechanisms responding to bacterial-derived molecules, oxidative stress, and reactive oxygen species. CIBERSORT statistics showed that both SNORD15A and RMRP were significantly positively correlated with CD4 memory-activated T cells, and plasma cells, and negatively correlated with T cell regulation. Mendelian randomization analysis showed a negative association between exercise and the risk of depression (OR: 0.965, p = 0.005). Conclusion: SNORD15A, RMRP in peripheral blood is negatively associated with the risk of MDD. The discovery of the signature exercise genes expands the understanding of the pathological mechanisms of MDD and opens up new perspectives on exercise therapy for MDD.