Major depressive disorder (MDD) is a prevalent and debilitating mental health condition characterized by persistent feelings of sadness and loss of interest. Despite its high prevalence, the underlying molecular mechanisms remain poorly understood. This study aims to elucidate the gene expression differences across distinct brain regions in MDD patients, identify potential diagnostic and therapeutic targets, and establish predictive models using bioinformatics approaches. Whole-transcriptome sequencing data from three different human brain regions were obtained from five datasets (GSE54564, GSE54571, GSE54572, GSE54567, GSE54568) in the GEO database. Gene symbol preprocessing was conducted using the XIANTAO platform. Differentially expressed genes (DEGs) were identified between MDD samples and controls using the R package "limma." Protein-protein interaction (PPI) networks were constructed using STRING and visualized in Cytoscape. Core genes were identified via CytoHubba using three algorithms (MCC, DEGREE, EPC). Receiver operating characteristic (ROC) curve analysis was performed to evaluate the prognostic value of core genes. LASSO regression was employed to enhance prediction accuracy and interpretability of machine learning models. Potential therapeutic drugs were predicted using the Comparative Toxicogenomics Database (CTD). In total, 342 DEGs related to the amygdala, 76 DEGs related to the anterior cingulate cortex, and 64 DEGs related to the dorsolateral prefrontal cortex were identified (p < 0.05, |logFC| > 0.15). Key diagnostic genes included COX5A and SST for the amygdala; CTSG, IL18RAP, LMO2, and MS4A7 for the anterior cingulate cortex; and VGF for the dorsolateral prefrontal cortex. The machine learning models demonstrated high predictive accuracy with AUC values of 0.776 for the amygdala, 0.928 for the anterior cingulate cortex, and 0.867 for the dorsolateral prefrontal cortex. Potential therapeutic drugs included dorsomorphin and trichostatin A. Gene set enrichment analysis (GSEA) revealed significant pathways such as oxidative phosphorylation in the amygdala, TYROBP microglial network in the anterior cingulate cortex, and MAPK signaling pathway in the dorsolateral prefrontal cortex. This study provides a comprehensive bioinformatics analysis of gene expression differences across brain regions in MDD patients. The identified core genes and pathways offer valuable insights into disease mechanisms and potential therapeutic targets, paving the way for future clinical applications.
