Recently, studies have reported a correlation that individuals with diabetes show an increased risk of developing Alzheimer’s disease (AD). Mulberry leaves, serving as both a traditional medicinal herb and a food source, exhibit significant hypoglycemic and antioxidative properties. The flavonoid compounds in mulberry leaf offer therapeutic effects for relieving diabetic symptoms and providing neuroprotection. However, the mechanisms of this effect have not been fully elucidated. This investigation aimed to investigate the combined effects of specific mulberry leaf flavonoids (kaempferol, quercetin, rhamnocitrin, tetramethoxyluteolin, and norartocarpetin) on both type 2 diabetes mellitus (T2DM) and AD. Additionally, the role of the gut microbiota in these two diseases’ treatment was studied. Using network pharmacology, we investigated the potential mechanisms of flavonoids in mulberry leaves, combined with gut microbiota, in combating AD and T2DM. In addition, we identified protein tyrosine phosphatase 1B (PTP1B) as a key target for kaempferol in these two diseases. Molecular docking and molecular dynamics simulations showed that kaempferol has the potential to inhibit PTP1B for indirect treatment of AD, which was proven by measuring the IC₅₀ of kaempferol (279.23 μM). The cell experiment also confirmed the dose-dependent effect of kaempferol on the phosphorylation of total cellular protein in HepG2 cells. This research supports the concept of food–medicine homology and broadens the range of medical treatments for diabetes and AD, highlighting the prospect of integrating traditional herbal remedies with modern medical research.

Reproductive toxicity poses significant risks to fertility and progeny health, making its identification in pharmaceutical compounds crucial. In this study, we conducted a comprehensive in silico investigation of reproductive toxic molecules, identifying three distinct categories represented by Dimethylhydantoin, Phenol, and Dicyclohexyl phthalate. Our analysis included physicochemical properties, target prediction, and KEGG and GO pathway analyses, revealing diverse and complex mechanisms of toxicity. Given the complexity of these mechanisms, traditional molecule-target research approaches proved insufficient. Support Vector Machines (SVMs) combined with molecular descriptors achieved an accuracy of 0.85 in the test dataset, while our custom deep learning model, integrating molecular SMILES and graphs, achieved an accuracy of 0.88 in the test dataset. These models effectively predicted reproductive toxicity, highlighting the potential of computational methods in pharmaceutical safety evaluation. Our study provides a robust framework for utilizing computational methods to enhance the safety evaluation of potential pharmaceutical compounds.