Exploring the therapeutic potential of Radix Puerariae isoflavonoids against atherosclerosis through integrative network pharmacology and machine learning.

Atherosclerosis (AS) is a chronic inflammatory disease of the vascular wall driven by a complex interplay between dysregulated immune responses and lipid retention. Its systemic complications remain the leading global cause of mortality. Despite current therapeutic strategies, substantial residual cardiovascular risk persists, largely due to insufficiently targeting of inflammatory and lipid-related pathways. Radix Puerariae (Gegen) is rich in isoflavonoids with reported cardioprotective properties; however, their coordinated anti-atherosclerotic mechanisms have not been fully characterized. Using an integrated network pharmacology framework, five active isoflavonoids (daidzein-4,7-diglucoside, formononetin, 3'-methoxydaidzein, puerarin, and 7,8,4'-trihydroxyisoflavonoid) were identified and predicted to target 79 AS-related proteins. Functional enrichment analyses indicated significant involvement in lipid metabolism, inflammatory response, apoptosis, and fluid shear stress pathways. Multi-algorithm machine learning (LASSO, SVM-RFE, RF, and NNET) consistently identified TNF and MMP9 as dominant diagnostic drivers, while RXRA emerged as a mechanistically relevant consensus target. Molecular docking and 100-ns molecular dynamics simulations indicated stable and favorable binding interactions between the isoflavonoids and their corresponding targets. Notably, daidzein-4,7-diglucoside consistently exhibited the strongest target-binding affinity, surpassing the well-studied puerarin, and is thus proposed as a potential novel lead candidate. Collectively, these findings suggest that the five Radix Puerariae isoflavonoids may exert coordinated anti-atherosclerotic effects through a multi-target, multi-pathway, and multicellular regulatory network, supporting their potential as a mechanism-driven adjunctive therapeutic strategy for AS.