• AI-chemometric assisted real-t...

Treffer: AI-chemometric assisted real-time monitoring of tryptophan fermentation process using a sensor fusion strategy.

Title:
AI-chemometric assisted real-time monitoring of tryptophan fermentation process using a sensor fusion strategy.
Authors:
Ma X; NMPA Key Laboratory for Technology Research and Evaluation of Drug Products, School of Pharmaceutical Sciences, Cheeloo College of Medicine, Shandong University, Jinan, Shandong 250012, China., Tian M; NMPA Key Laboratory for Technology Research and Evaluation of Drug Products, School of Pharmaceutical Sciences, Cheeloo College of Medicine, Shandong University, Jinan, Shandong 250012, China., Huang R; NMPA Key Laboratory for Technology Research and Evaluation of Drug Products, School of Pharmaceutical Sciences, Cheeloo College of Medicine, Shandong University, Jinan, Shandong 250012, China., Zhang H; National Glycoengineering Research Center, Shandong University, Qingdao, Shandong 266200, China., Yue J; NMPA Key Laboratory for Technology Research and Evaluation of Drug Products, School of Pharmaceutical Sciences, Cheeloo College of Medicine, Shandong University, Jinan, Shandong 250012, China., Jia H; Channel Biotech Co., Ltd., Innovation R&D Center, Jinan, Shandong 250012, China., Wang Y; Channel Biotech Co., Ltd., Innovation R&D Center, Jinan, Shandong 250012, China., Li Y; Single-Cell Center, Key Laboratory of Photoelectric Conversion and Utilization of Solar Energy, Qingdao Institute of BioEnergy and Bioprocess Technology, Chinese Academy of Sciences, Qingdao 266101, China., Gai C; Dongying XINGSHENG Special Equipment Technology Co., Ltd., Dongying, Shandong 257000, China., Li L; NMPA Key Laboratory for Technology Research and Evaluation of Drug Products, School of Pharmaceutical Sciences, Cheeloo College of Medicine, Shandong University, Jinan, Shandong 250012, China; Key Laboratory of Chemical Biology (Ministry of Education), Shandong University, Jinan 250012, China. Electronic address: lilian@sdu.edu.cn., Zang H; NMPA Key Laboratory for Technology Research and Evaluation of Drug Products, School of Pharmaceutical Sciences, Cheeloo College of Medicine, Shandong University, Jinan, Shandong 250012, China; Key Laboratory of Chemical Biology (Ministry of Education), Shandong University, Jinan 250012, China. Electronic address: zanghcw@126.com.
Source:
Food chemistry [Food Chem] 2025 Oct 15; Vol. 489, pp. 145037. Date of Electronic Publication: 2025 Jun 03.
Publication Type:
Journal Article; Evaluation Study
Language:
English
Journal Info:
Publisher: Elsevier Applied Science Publishers Country of Publication: England NLM ID: 7702639 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1873-7072 (Electronic) Linking ISSN: 03088146 NLM ISO Abbreviation: Food Chem Subsets: MEDLINE
Imprint Name(s):
Publication: Barking : Elsevier Applied Science Publishers
Original Publication: Barking, Eng., Applied Science Publishers.
MeSH Terms:
Tryptophan*/metabolism , Tryptophan*/analysis , Chemometrics*/methods , Bacteria*/metabolism, Fermentation ; Machine Learning
Contributed Indexing:
Keywords: Fermentation; In-line monitoring; Machine learning; Multi-source sensor; NIR; Tryptophan
Substance Nomenclature:
8DUH1N11BX (Tryptophan)
Entry Date(s):
Date Created: 20250610 Date Completed: 20250715 Latest Revision: 20250715
Update Code:
20250715
DOI:
10.1016/j.foodchem.2025.145037
PMID:
40494072
Database:
MEDLINE

Weitere Informationen

Tryptophan, an essential amino acid, crucially impacts neuronal function, metabolism, immunity, and gut homeostasis. Microbial fermentation is the mainstream method for tryptophan production. The precise production process is essential for ensuring both high quality and optimal yield. This study aims to utilize AI-chemometrics methods to achieve the integrated monitoring of multi-source sensors. First, machine learning methods were applied to build in-line NIR prediction models. Then, a sensor fusion strategy was introduced to established the multivariate statistical process control (MSPC) model based on five in-line sensor data. The results showed that Gaussian process regression models were best for bacterial optical density, residual sugar, and tryptophan concentration. The validation sets RPD were 5.686, 3.297, and 3.130, respectively. MSPC charts synergistic analysis based on feature-level fusion enables real-time simultaneous detection of multi-source anomalies. This study provides an effective quality control strategy for food fermentation process to ensure consistent, stable and controllable product quality.
(Copyright © 2025. Published by Elsevier Ltd.)

Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.