Identification of Hub Genes Associated With Oxidative Stress of Acute Kidney Injury via Machine Learning Algorithm

Acute kidney injury (AKI) is a short-term clinical condition that results in a decline in renal function. The pathophysiology of AKI was influenced by a variety of pathways, one of which is oxidative stress (OS). The development of AKI was aided by increased OS from reactive oxygen species (ROS) and inadequate antioxidant mechanisms. Effective early biomarkers and treatment approaches for AKI are still lacking; however, the widespread use of bioinformatic analysis may help.

In order to find co-expressed genes, we used three different machine learning techniques to evaluate the Gene Expression Omnibus (GEO) collection of AKI and genes related to oxidative stress (OS) from the Gene Ontology (GO) database. GSEA and DAVID were used to analyze the related pathways. The String database was used to build a protein-protein interaction (PPI) network, and HK-2 cells were used to create an in vitro hypoxia-reoxygenation (H/R) model. Additionally, its clinical utility was further validated by machine learning and the pan-cancer study of the intersection gene (HGF) in PPI.

Thus, AK4, PLA2R1, HGF, VNN1, PPARGC1A, and VKORC1L1 were hub genes that may be important regulatory factors and biomarkers of OS in AKI, and 36 OS-DEGs were found. The gene expression level and prediction effectiveness were confirmed by the ROC curve and RT-qPCR data, respectively.

The molecular mechanism of OS in AKI is now better understood because to these studies, which also offer fresh perspectives on AKI research and therapy.