Exploring the molecular mechanisms of Pueraria in Alzheimer’s disease treatment using machine learning and network pharmacology

Background Alzheimer’s disease (AD) is a multifactorial neurodegenerative disorder, characterized by amyloid- β deposition, tau pathology, neuroinflammation, and metabolic dysfunction. While conventional treatments have been widely studied, food-based interventions are emerging as potential neuroprotective strategies. Pueraria, a nutrient-rich food, has shown promise in promoting brain health, but its mechanisms in AD prevention and management remain insufficiently understood. Methods In this study, we utilized network pharmacology, transcriptomics, and machine learning to investigate the neuroprotective effects of Pueraria. Through analysis of five transcriptomic datasets (GSE5281, GSE29378, GSE36980, GSE37263, and GSE138260), we identified genes associated with AD and screened 15 active compounds from Pueraria lobata using HERB and TCMSP databases. Machine learning models prioritized key targets, and molecular docking simulations assessed the binding affinities of Pueraria compounds to these targets. In vivo validation was performed in AD model mice to evaluate the cognitive-enhancing effects of Pueraria. Results We identified 45 overlapping targets between Pueraria and AD, primarily related to synaptic plasticity and neurotransmission. Among these, PFKFB3 emerged as a key mediator of Pueraria’s neuroprotective effects. Molecular docking confirmed strong binding affinities between Pueraria compounds and PFKFB3, supporting their functional role. Experimental data showed that Pueraria improved cognitive function in AD mice, underscoring its potential as a neuroprotective agent. Conclusion This study highlights Pueraria as a promising functional food for AD prevention and management, emphasizing the potential of plant-based dietary interventions for brain health. Our findings provide a basis for further exploration of food-derived neuroprotective strategies.