领学术科研之先,创食品科技之新
—— 中国食品杂志社
期刊集群
A review of deep learning models for food flavor data analysis
来源:导入 阅读量: 2 发表时间: 2026-02-10
作者: Jiawen Yuan, Qihan Wu, Jie Zhou, Shuai Yu, Xing Xin, Jin Liu, Xiaohui Cui
关键词: Umami peptides;Virtual enzymatic;Virtual screening;biosensor
摘要:

a non-experimental approach based on protein sequences of Takifugu rubripes (a well-known fish for its umami taste) was initially employed using the Procleave database for virtual endogenous enzyme digestion, aimed at screening and identifying novel umami peptides. the potential umami peptides were screened through a dual-strategy approach combining sequence-based analysis and molecular docking techniques. Consequently, five umami peptides (faGDDaPr, HEGEQGQEG, aaPHENatLH, EsHQQtLDD, and GEVED) were selected, synthesized and subjected to comprehensive sensory evaluation, electronic tongue assessment and kinetic analysis using the t1r1-Vft biosensor. sensory evaluation demonstrated that these peptides exhibited low umami thresholds, ranging from 0.195 to 0.281 mg/mL, with primary binding sites at asn150, ser248, arg151, thr154, and Lys155 on the receptor. the kinetic analysis revealed a linear correlation between the kinetic values and peptide concentrations within the range of 10–15–10–5 mg/mL, consistent with sensory evaluation results. therefore, the virtual hydrolysis and screening strategies, designed for the rapid identification of novel umami peptides from T. rubripes, emerge as a practicable and efficient means for the rapid acquisition of peptides and the enrichment of the umami peptides database.

电话: 010-87293157 地址: 北京市丰台区洋桥70号

版权所有 @ 2023 中国食品杂志社 京公网安备11010602060050号 京ICP备14033398号-2