An integrated framework combining spectral prediction based on amino acid profiling and network modularity analysis:...

Boletus bainiugan, valued both as a food and a research subject, is rich in various functional components. Its amino acid profile provides a key entry point for exploring its potential functionality. To evaluate and uncover the functional potential of Boletus bainiugan, this study constructed an integrated framework combining spectral prediction and network modularity analysis. The amino acid profiles of five Boletus species and Boletus bainiugan under different drying temperatures were analyzed using UPLC-MS, leading to the identification of seven amino acids and derivatives (AADs) as quality markers. A target-disease network was constructed, and weighted Leiden algorithm-based network modularity analysis identified five key functional communities. Molecular docking confirmed stable binding (from -4 to -6.5 kcal/mol) between key AADs and their targets. Furthermore, a multi-output prediction model based on FT-IR was established. The R² value of the model exceeded 0.96, enabling rapid and non-destructive prediction of the contents of key AADs. This study provides an efficient, reliable, and high-throughput methodological framework for functional exploration and quality control of natural products, contributing to further research in this field.