Combining bioinformatics and network pharmacology to explore diagnostic biomarkers and mechanisms associated with mitochondria in diabetic nephropathy using Rhodiola Rosea

Abstract This study aimed to explore diagnostic biomarkers associated with mitochondria in the treatment of diabetic nephropathy (DN) using Rhodiola Rosea (RR) and to elucidate the underlying regulatory mechanisms. The target genes corresponding to the active ingredients of RR, mitochondria-related genes, and DN-related genes were intersected with differentially expressed genes from GSE96804 to obtain mitochondria-related pharmacopoeia genes (MRPGs). Based on these genes, a pharmacological network was constructed. Additionally, the biological pathways enriched by the diagnostic biomarkers, their correlation with the DN immune infiltration profile, and their binding ability to RR were analyzed. A total of 24 MRPGs against DN were identified using Venn analysis. Subsequently, six diagnostic biomarkers were screened by machine learning, namely CAT, CASP3, CPT1A, OGDH, CYP27B1, and ALDH9A1. Immune infiltration analysis showed eight differential immune cells between the DN and control groups, with macrophages M2 and neutrophils correlating with all diagnostic biomarkers. Based on the pharmacological network and molecular docking, two targeting relationship pairs were identified: CAT-RUTIN and ALDH9A1-RUTIN-2, both demonstrating good binding capacity. This study identified six diagnostic biomarkers for DN based on the effects of RR and mitochondrial function. The findings revealed the active components and potential molecular therapeutic mechanisms of RR in treating DN.