Minhao Xu , Hongjin Ning , Zhijun Wu , Hong Chen , Wen Qin , Derong Lin
{"title":"Exploring the modification of dietary fiber from black highland barley bran based on machine learning and metabolomics","authors":"Minhao Xu , Hongjin Ning , Zhijun Wu , Hong Chen , Wen Qin , Derong Lin","doi":"10.1016/j.jcs.2025.104277","DOIUrl":null,"url":null,"abstract":"<div><div>This study systematically investigated the impact of different treatment parameters on dietary fiber components and their underlying modification mechanisms concerning functional properties through machine learning and metabolomics. A Random Forest (RF) model was established to conduct a comprehensive analysis on multi-parameter optimization, achieving superior predictive performance (test set R<sup>2</sup> = 0.9602) and identifying optimal soluble dietary fiber (SDF) production conditions: microwave power at 550 W, microwave duration at 3 min, inoculation volume at 11 %, fermentation temperature at 36 °C, and fermentation time at 30 h. Metabolomic profiling revealed that organic acids produced during microbial fermentation altered the structure of insoluble dietary fiber (IDF), thereby affecting its physicochemical properties. This study successfully constructed a mathematical relationship model between dietary fiber components and functional characteristics, providing mechanistic insights into dietary fiber modification processes.</div></div>","PeriodicalId":15285,"journal":{"name":"Journal of Cereal Science","volume":"126 ","pages":"Article 104277"},"PeriodicalIF":3.7000,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Cereal Science","FirstCategoryId":"97","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0733521025001766","RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"FOOD SCIENCE & TECHNOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
This study systematically investigated the impact of different treatment parameters on dietary fiber components and their underlying modification mechanisms concerning functional properties through machine learning and metabolomics. A Random Forest (RF) model was established to conduct a comprehensive analysis on multi-parameter optimization, achieving superior predictive performance (test set R2 = 0.9602) and identifying optimal soluble dietary fiber (SDF) production conditions: microwave power at 550 W, microwave duration at 3 min, inoculation volume at 11 %, fermentation temperature at 36 °C, and fermentation time at 30 h. Metabolomic profiling revealed that organic acids produced during microbial fermentation altered the structure of insoluble dietary fiber (IDF), thereby affecting its physicochemical properties. This study successfully constructed a mathematical relationship model between dietary fiber components and functional characteristics, providing mechanistic insights into dietary fiber modification processes.
期刊介绍:
The Journal of Cereal Science was established in 1983 to provide an International forum for the publication of original research papers of high standing covering all aspects of cereal science related to the functional and nutritional quality of cereal grains (true cereals - members of the Poaceae family and starchy pseudocereals - members of the Amaranthaceae, Chenopodiaceae and Polygonaceae families) and their products, in relation to the cereals used. The journal also publishes concise and critical review articles appraising the status and future directions of specific areas of cereal science and short communications that present news of important advances in research. The journal aims at topicality and at providing comprehensive coverage of progress in the field.