Weighted gene co-expression network analysis and machine learning identified the lipid metabolism-related gene LGMN as a novel biomarker for keloid.

Abstract

The aetiology of keloid formation remains unclear, and existing treatment modalities have not definitively established a successful approach. Therefore, it is necessary to identify reliable and novel keloid biomarkers as potential targets for therapeutic interventions. In this study, we performed differential expression analysis and functional enrichment analysis on the keloid related datasets, and found that multiple metabolism-related pathways were associated with keloid formation. Subsequently, the differentially expressed genes (DEGs) were intersected with the results of weighted gene co-expression network analysis (WGCNA) and the lipid metabolism-related genes (LMGs). Then, three learning machine algorithms (SVM-RFE, LASSO and Random Forest) together identified legumain (LGMN) as the most critical LMGs. LGMN was overexpressed in keloid and had a high diagnostic performance. The protein-protein interaction (PPI) network related to LGMN was constructed by GeneMANIA database. Functional analysis of indicated PPI network was involved in multiple immune response-related biological processes. Furthermore, immune infiltration analysis was conducted using the CIBERSORT method. M2-type macrophages were highly infiltrated in keloid tissues and were found to be significantly and positively correlated with LGMN expression. Gene set variation analysis (GSVA) indicated that LGMN may be related to promoting fibroblast proliferation and inhibiting their apoptosis. Moreover, eight potential drug candidates for keloid treatment were predicted by the DSigDB database. Western blot, qRT-PCR and immunohistochemistry staining results confirmed that LGMN was highly expressed in keloid. Collectively, our findings may identify a new biomarker and therapeutic target for keloid and contribute to the understanding of the potential pathogenesis of keloid.