主成分分析
代谢组学
色谱法
四极飞行时间
化学
质谱法
线性判别分析
化学计量学
层次聚类
模式识别(心理学)
聚类分析
数学
人工智能
统计
计算机科学
串联质谱法
作者
Lei Xu,Fei Lao,Zhenzhen Xu,Xue Wang,Fang Chen,Xiaojun Liao,Ailiang Chen,Shuming Yang
标识
DOI:10.1016/j.foodchem.2019.01.154
摘要
An untargeted metabolomic method based on UPLC-QTOF were used to investigate the differences in coffee brewed by boiled, pour-over and cold-brew methods here. Distinctive separation among the three groups could be seen from principal component analysis and hierarchical clustering analysis. Analysis of variance, fold change and orthogonal projection to latent structures discriminant mode were conducted to find the characteristic potential markers, subsequently, nine potential markers were putatively identified using general chemical databases, and five of them were further confirmed by acquisition of reference standards. This work provides an efficient way for discrimination of coffee brewed by different methods. Interestingly, the result of this work also suggested that the contents of two selected markers, norharman and harman, were higher in the pour-over and boiled methods, compared to the cold-brew method. This content difference were further verified by the quantitative analysis data of commercial coffee samples.
科研通智能强力驱动
Strongly Powered by AbleSci AI