Fig 1.
Identification of DEGs in MDD and vitiligo.
(A, B) Volcano map of DEGs in MDD (A) and vitiligo (B). (C, D) Heatmaps of hierarchical clustering of the top 25 DEGs in MDD (C) and vitiligo (D). (E) Identification of 156 co-expressed genes by overlapping the DEGs in MDD and vitiligo.
Fig 2.
Screening hub genes and functional enrichment analysis of hub genes.
(A) Protein-protein interaction network of common DEGs in MDD and vitiligo. (B) The 14 hub genes with the highest connectivity based on the CytoHubba-MCC algorithm. (C) GO and KEGG enrichment analysis of 14 hub genes.
Fig 3.
Identification of key genes by different machine learning algorithms.
(A, D) Feature selection using the LASSO regression method in MDD (A) and vitiligo (D). (B, E) Feature selection using the SVM‑RFE in MDD (B) and vitiligo (E). (C, F) Gene importance ranking derived from random forest analysis in MDD (C) and vitiligo (F). (G) UpSet diagram illustrating the intersection of genes identified by the three machine‑learning algorithms.
Fig 4.
ssGSEA analysis of immune infiltration.
(A, B) Boxplot of immune cell enrichment scores inferred by ssGSEA in MDD (A) and vitiligo (B). (C, D) The correlation heatmaps between immune cell enrichment scores in MDD (C) and vitiligo (D). (E, F) The correlation between key genes (EXOSC7, KLRG1, and MAPK14) and immune cell enrichment scores in MDD. (G, H) The correlation between key genes (EXOSC7, KLRG1, and MAPK14) and immune cell enrichment scores in vitiligo. * P-value < 0.05, ** P-value < 0.01, *** P-value < 0.001.
Fig 5.
The interaction network of key genes.
(A) The interaction network of key genes via GeneMANIA. (B) Co-regulatory network of TF-miRNA and three key genes.
Fig 6.
Validation of key genes and single‑gene GSEA analysis.
(A, B) Validation of the expression of key genes in the MDD dataset (A) and vitiligo dataset (B) based on T-test. * P-value < 0.05, ** P-value < 0.01, *** P-value < 0.001. (C, D) ROC curves of key genes to assess their discriminatory ability in MDD (C) and vitiligo (D). (E, F) ROC curves of key genes to evaluate their performance in the external validation datasets GSE52790 (E) and GSE80009 (F). (G, H) Representative enriched pathways identified by single‑gene GSEA for MAPK14 in MDD (G) and vitiligo (H).
Table 1.
Prediction of key genes-Related Drugs.