Du Lijuan, Lu Weiying, Zhang Yaqiong, Gao Boyan, Yu Liangli
Department of Food Science and Technology School of Agriculture and Biology Institute of Food and Nutraceutical Science Shanghai Jiao Tong University Shanghai China.
China-Canada Joint Lab of Food Nutrition and Health (Beijing) Beijing Technology & Business University (BTBU) Beijing China.
Food Sci Nutr. 2020 Feb 3;8(3):1471-1479. doi: 10.1002/fsn3.1430. eCollection 2020 Mar.
Detection of the presence of milk powder in liquid whole milk is challenging due to their similar chemical components. In this study, a sensitive and robust approach has been developed and tested for potential utilization in discriminating adulterated milk from liquid whole milk by analyzing the intact protein and hydrolyzed peptide using ultra-performance liquid chromatography with quadrupole time-of-flight mass spectrometer (UPLC-QTOF-MS) fingerprints combined with data fusion. Two different datasets from intact protein and peptide fingerprints were fused to improve the discriminating ability of principle component analysis (PCA). Furthermore, the midlevel data fusion coupled with PCA could completely distinguish liquid whole milk from the milk. The limit of detection of milk powder in liquid whole milk was 0.5% (based on the total protein equivalence). These results suggested that fused data from intact protein and peptide fingerprints created greater synergic effect in detecting milk quality, and the combination of data fusion and PCA analysis could be used for the detection of adulterated milk.
由于全脂液态奶和奶粉的化学成分相似，检测全脂液态奶中是否存在奶粉具有挑战性。在本研究中，开发并测试了一种灵敏且可靠的方法，通过使用超高效液相色谱-四极杆飞行时间质谱仪（UPLC-QTOF-MS）指纹图谱结合数据融合技术分析完整蛋白质和水解肽，以鉴别全脂液态奶中的掺假牛奶。将来自完整蛋白质和肽指纹图谱的两个不同数据集进行融合，以提高主成分分析（PCA）的鉴别能力。此外，中级数据融合与PCA相结合能够完全区分全脂液态奶和掺假牛奶。全脂液态奶中奶粉的检测限为0.5%（基于总蛋白质当量）。这些结果表明，完整蛋白质和肽指纹图谱的融合数据在检测牛奶质量方面产生了更大的协同效应，并且数据融合与PCA分析相结合可用于检测掺假牛奶。
