Integrating network pharmacology, molecular docking and simulation approaches with machine learning reveals the multi-target pharmacological mechanism of Berberis integerrima against diabetic nephropathy

Diabetic nephropathy (DN) is one of the most feared complications of diabetes and key cause of end-stage renal disease (ESRD). Berberis integerrima has been widely used to treat diabetic complications, but exact molecular mechanism is yet to be discovered. Data on active ingredients of B. integerrima and target genes of both diabetic nephropathy and B.integerrima were obtained from public databases. Common results between B. integerrima and DN targets were used to create protein-protein interaction (PPI) network using STRING database and exported to Cytoscape software for the selection of hub genes based on degree of connectivity. Future, PPI network between constituents and overlapping targets was created using Cytoscape to investigate the network pharmacological effects of B. integerrima on DN. KEGG pathway analysis of core genes exposed their involvement in excess glucose-activated signaling pathway. Then, expression of core genes was validated through machine learning classifiers. Finally, PyRx and AMBER18 software was used for molecular docking and simulation. We found that Armepavine, Berberine, Glaucine, Magnoflorine, Reticuline, Quercetin inhibits the growth of diabetic nephropathy by affecting ICAM1, PRKCB, IKBKB, KDR, ALOX5, VCAM1, SYK, TBXA2R, LCK, and F3 genes. Machine learning revealed SYK and PRKCB as potential genes that could use as diagnostic biomarkers against DN. Furthermore, docking and simulation analysis showed the binding affinity and stability of the active compound with target genes. Our study revealed that B. integerrima has preventive effect on DN by acting on glucose-activated signaling pathways. However, experimental studies are needed to reveal biosafety profiles of B. integerrima in DN.