Identifying MRPS10 as a Diagnostic Biomarker for MDD via Machine Learning

Major depressive disorder (MDD) is a serious psychological disorder, which can lead to a high disability and mortality rate. Two GEO datasets were merged to create the analysis dataset. A total of 17,598 differentially expressed genes (DEGs) were utilized for Gene Set Enrichment Analysis (GSEA) to identify pathways distinguishing the MDD group from the control group. Notably, the gene MRPS10 emerged as a prominent candidate for a diagnostic biomarker of MDD, particularly indicating the late-stage condition, as identified by Least Absolute Shrinkage and Selection Operator (LASSO) and Support Vector Machine Recursive Feature Elimination (SVM-RFE). Receiver Operating Characteristic (ROC) curves for MRPS10 were employed to demonstrate its discriminatory power. The CIBERSORT algorithm was utilized to evaluate the distribution of tissue-infiltrating immune cells in both the MDD and control groups. The diagnostic biomarker, the gene MRPS10, demonstrated a positive correlation with resting dendritic cells and M1 macrophages, and a negative correlation with monocytes and activated NK cells. This underscores the significant role of this gene in immune cell infiltration. To bolster the reliability of our findings, we conducted an additional analysis on the expression of the gene MRPS10 within single-cell transcriptome data, which revealed significant differences in expression levels between the main excitatory neuron (EX) and inhibitory neuron (IN) types EX/L2/4, EX/L4/6, and IN/VIP. In summary, our research has pinpointed the gene MRPS10 as a biomarker, thereby enhancing the knowledge repository pertinent to clinical diagnostics and pharmaceutical development for MDD.