Integrating Multi-Omics Data Using Machine Learning to Explore New Therapeutic Targets for Acute Kidney Injury

Renal ischemia–reperfusion (I/R) injury is an unavoidable complication associated with renal transplantation, and currently, there are no targeted therapeutic interventions. The objective of this study was to explore the molecular mechanisms that contribute to I/R-induced acute kidney injury (I/R-AKI) and to discover potential targets for effective renal safeguarding. Bioinformatics techniques were employed to analyse critical genes regulating I/R-AKI at the single-cell level and to develop diagnostic models. Additionally, key pharmacological agents that inhibit the expression of target genes were identified for subsequent experimental validation. Pathological changes in the kidneys of I/R mice and patients with AKI were observed using immunofluorescence, western blotting, immunohistochemistry and transmission electron microscopy. We developed and validated a robust diagnostic model for I/R-AKI. The results suggest that ADAMTS1 acts as a promoter of renal I/R-AKI. In I/R-AKI, ADAMTS1 was significantly upregulated in renal tubular epithelial cells. Furthermore, apoptosis mediated by the mitochondrial pathway was a critical factor in the progression of renal I/R injury. In mouse models of I/R-AKI, the inhibition of ADAMTS1 with troglitazone significantly reduced both functional and histological damage. The diagnostic model can serve as a valuable instrument for diagnosing I/R-AKI. Furthermore, troglitazone can significantly contribute to managing I/R-AKI by inhibiting the expression of ADAMTS1. This study provides critical insights that may inform future research on therapeutic targets for renal ischaemia–reperfusion injury.