An analysis of gene expression profiles through machine learning uncovers the new diagnostic signature for diabetic foot ulcers.

Purpose: Diabetic foot ulcers (DFUs), a serious diabetes complication, greatly increase disability and mortality, underscoring the need for effective diagnostic markers.

Methods: We used GSE199939 and GSE134431 datasets from the Gene Expression Omnibus (GEO) database, removed batch effects, and identified differentially expressed genes (DEGs). The weighted gene co-expression network analysis (WGCNA) was used to identify co-expression modules, followed by the integration of the protein-protein interaction (PPI) network to screen key genes, which were further optimized using LASSO regression. The gene set enrichment analysis (GSEA) analyzed key gene-related pathways, CIBERSORT assessed immune infiltration, and potential target drugs were predicted using the DGIdb database.

Results: We identified 403 DEGs in DFUs, intersected them with 2,342 genes from a DFU-related WGCNA module to find 193 overlapping genes, and screened candidates via PPI network. LASSO regression finalized DCT, PMEL, and KIT as the key genes. GSEA analysis showed these three genes may influence the MAPK and PI3K-Akt pathways and were positively correlated with Dendritic. cells.resting. Drug target prediction identified 85 potential drugs for KIT, six for DCT, and six for PMEL.

Conclusion: This research highlights DCT, PMEL, and KIT as diagnostic biomarkers for DFUs, which are linked to melanin production and the MAPK/PI3K-Akt signaling pathways.