Identification and characterization of umami peptides from fermented soybean meal hydrolysates: Combined machine learning, cellular functional characterization, and structural interpretation.

This study developed an efficient strategy for identifying umami and umami-enhancing peptides from fermented soybean meal hydrolysates (FSMH) to support the high-value utilization of soybean byproducts. An integrated approach combining virtual screening, molecular docking, sensory validation, intracellular calcium mobilization assays, and molecular dynamics simulation was employed. Sixteen peptides were synthesized for sensory validation, confirming 15 umami peptides. LE, AE, GEDLMVQ, and FEEINKV exhibited the strongest enhancement in both monosodium glutamate and inosinate/disodium guanylate systems, with GEDLMVQ showing the lowest enhancement threshold (0.0079 mM). Fermentation significantly increased most umami peptide concentrations, peaking at 24 h. Molecular docking implicated T1R1-VFTD as the main binding domain, mediated primarily by hydrogen bonding, hydrophobic contacts, and van der Waals forces. Representative peptides induced TAS1R1-associated Ca2+ responses with distinct kinetics. Molecular dynamics simulations provided supportive structural information for interpreting peptide-receptor interactions. These findings support a feasible strategy for obtaining sustainable flavor enhancers from soybean byproducts.