Development and Validation of a Diagnostic Model for AKI Based on the Analysis of Ferroptosis-related Genes.

Abstract

Acute kidney injury (AKI) is a common renal condition associated with various factors, including pre-renal, post-renal, and renal causes, with ischemia- reperfusion being a frequent contributor leading to tubular injury. Early identification of AKI is crucial but remains challenging. This study explored the molecular signature of AKI using gene microarray data from the GEO dataset, focusing on identifying ferroptosis-related features through three machine-learning algorithms. We also validated potential biomarkers through a hypoxia/ reoxygenation model. ROC curves, expression differences, and associations with immune cells were analyzed for the three markers to confirm their potential as AKI biomarkers, each demonstrating strong diagnostic ability. Combining these markers proved more effective. The combination of AEBP2, MDM2, and NR4A1 as diagnostic biomarkers for AKI not only enhances detection capability but also holds promise as a significant tool in clinical practice, providing patients with diagnostic and therapeutic guidance.