Identification of signature genes for renal ischemia‒reperfusion injury based on machine learning and WGCNA

• DOI: 10.1016/j.heliyon.2023.e21151
• Publication Date: 2023-10-18T08:04:53Z
• Authors: Zechao Lu; Senkai Xu; Haiqin Liao; Yixin Zhang; Zeguang Lu; Zhibiao Li; Yushu Chen; Feng Guo; Fucai Tang; Zhaohui He
• Keywords: Candidate gene; Lasso; Signature (topology)

As an inevitable event after kidney transplantation, ischemia‒reperfusion injury (IRI) can lead to a decrease in transplant success. The search for signature genes of renal (RIRI) is helpful improving the diagnosis and guiding clinical treatment.We first downloaded 3 datasets from GEO database. Then, differentially expressed (DEGs) were identified applied functional enrichment analysis. After that, we performed three machine learning methods, including random forest (RF), Lasso regression analysis, support vector recursive feature elimination (SVM-RFE), further predict candidate genes. WGCNA was also executed screen DEGs. took intersection obtain RIRI. Receiver operating characteristic (ROC) analysis conducted measure predictive ability Kaplan‒Meier used association between graft survival. Verifying expression ischemia cell model.A total 117 DEGs screened out. Subsequently, RF, SVM-RFE 17, 25, 18 74 genes, respectively. Finally, (DUSP1, FOS, JUN) out through ROC suggested that could well diagnose indicated patients with low FOS or JUN had longer OS than those high expression. validated using model compared control group, level increased under hypoxic conditions.Three offer good prediction RIRI outcome may serve as potential therapeutic targets intervention, especially JUN. survival by improve