In silico analysis of soybean-derived umami peptides: Discovery and interaction mechanisms with T1R1/T1R3 receptor.

In this study, umami peptides with binding activity to the umami receptor T1R1/T1R3 were screened and identified from soybean protein. Using virtual enzymatic hydrolysis, a total of 629 dipeptides to hexapeptides were generated. Through predictions of bioactivity, water solubility, and hemolytic activity, 43 non-toxic peptides were selected. Deep learning methods were employed to predict the umami characteristics of these peptides, ultimately leading to the identification of 17 peptides with potential umami properties. Further molecular docking analysis revealed that the peptides DSWPSL, SHHPR, LGPK and SSW exhibited high binding stability with the umami receptor, indicating strong umami characteristics. The umami properties of these peptides were confirmed through electronic tongue experiments and sensory evaluation, with SHHPR exhibiting the lowest bitterness in sensory evaluation, making it seem more suitable for consumption in food. Molecular dynamics simulations uncovered the interaction mechanisms between the umami peptides and T1R1/T1R3, highlighting charge-charge forces as the primary interaction. This study not only provides new insights for the development of natural umami enhancers but also demonstrates the integration of food science and computational techniques.