基于GC-IMS技术的油菜花蜂蜜产地溯源模型鉴别

Identification Model for Rape Flower Honey Geographical Origin Based on GC-IMS Technology

采用气相离子迁移谱(GC-IMS)测定全国10个产地油菜花蜂蜜的挥发性有机化合物(VOCs),通过二维差谱法从广泛的VOCs信号数据中筛选出67个有效表征油菜花蜂蜜产地差异的特征成分,从差异化的VOCs分布轮廓的非靶向角度对10个产地油菜花蜂蜜进行主成分分析(PCA)和偏最小二乘判别分析(PLS-DA),并基于PLS-DA构建油菜花蜂蜜产地溯源鉴别模型。结果表明:交叉验证中模型的预测能力参数(Q^(2))和拟合优度参数(R^(2))分别为0.959和0.987,置换模拟验证未发生过拟合情况,模型稳健有效,可实现对10个产地的油菜花蜂蜜的有效鉴别。同时,通过变量投影重要性(VIP)值进一步筛选出21个对有效区分油菜花蜂蜜产地发挥关键作用的特征标志物。本研究引入GC-IMS技术和PLS-DA的理论与方法实现了不同产地蜂蜜样品的准确鉴别,为蜂蜜的产地溯源提供了新的技术策略。

The gas chromatography-ion mobility spectrometry(GC-IMS)was used to determine the volatile organic compounds(VOCs)of rape flower honey from 10 geographical origin in China,67 characteristic components were selected from a wide range of VOCs signal data by two-dimensional differential spectrum method.Principal component analysis(PCA)and partial least squares discriminant analysis(PLS-DA)were performed on 10 geographical origin rape flower honey samples from the untargeted perspective of differentiated VOCs distribution profile.Based on PLS-DA,the discrimination model of rape flower honey geographical origin was established.The result showed that the predictive power(Q^(2))and goodness of fit(R^(2))parameters of the cross validation were 0.959 and 0.987,respectively.The permutation showed that it did not overfitting,and the identification model was robust and effective,which could effectively identify rape flower honey from 10 geographical origin.At the same time,21 characteristic markers that played a key role in effectively distinguishing rape flower honey geographical origin were further screened out by variable importance for the projection(VIP)values.In this study,GC-IMS technology and PLS-DA theory and method were introduced to achieve accurate identification of rape flower honey samples from different geographical origin,which provided a new technical strategy for honey origin tracing.