基于代谢组学的食用植物油真实性鉴别方法研究

Adulteration of edible vegetable is a challenge that consumers and researchers are facing. Traditional DNA markers are not stable at high temperatures and cannot withstand chemical treatments used during food processing. Therefore, alternative markers, such as endogenous metabolites, need to be explored for authentication of processed food. In this project, the metabolomic methods are established for phenylpropanoid derivatives and terpenoids in oilseeds and edible vegetable oils by lipid chromatography high resolution mass spectrometry (LC-MS). Moreover, the cloud database of characteristic components in oilseeds and edible oils will be built. More importantly, the core of these studies is propose the new algorithms such as one-class classification, variable selection, model evaluation and application domain prediction, and time series analysis. The strategy of "the common components for authenticity, the characteristic markers for adulteration" was proposed and employed to build adulteration detection methods for edible oils. The completion of the project will not only provide theoretical and technical support for the adulteration detection for edible vegetable oils, but also provide good theoretical support and reference for other food authentication.

食用植物油掺假严重制约产业发展。传统上食品掺假识别标志物DNA在油脂加工中易断裂，难以应用于食用油真伪鉴别，食用植物油掺假鉴别是长期困扰食品安全与质量领域的国际性难题。本项目重点针对真实性鉴别模型预测准确度提升和高特异性高普适性标志物筛查的难题，拟发展化学计量学变量选择、关系型大数据处理方法，建立食用油真实性鉴别模型；构建我国油料油脂特征组分云数据库，发展基于代谢网络分析的高特异性高普适性标志物筛查方法，建立化学计量学模型与特征标志物结合的食用油真实性精准鉴别技术，建立“共有组分保真，特征成分鉴伪”的食用植物油真实性鉴别技术体系。本项目的顺利完成将为食用植物油真伪鉴别提供直接的理论和技术支撑，为其它食品真伪鉴别提供良好的理论支撑和借鉴。

食用植物油掺假严重制约产业发展。传统上食品掺假识别标志物DNA在油脂加工中易断裂，难以应用于食用油真伪鉴别，食用植物油掺假鉴别是长期困扰食品安全与质量领域的国际性难题。本项目重点针对真实性鉴别模型预测准确度提升和高特异性高普适性标志物筛查的难题，发展了化学计量学变量选择、关系型大数据处理方法，建立食用油真实性鉴别模型；构建了我国油料油脂特征组分云数据库，发展基于代谢网络分析的高特异性高普适性标志物筛查方法，建立了化学计量学模型与特征标志物结合的食用油真实性精准鉴别技术，发表论文23篇，其中SCI论文13篇，获授权专利4件，其中国际专利3件，论文建成了“共有组分保真，特征成分鉴伪”的食用植物油真实性鉴别技术体系。本项目的顺利完成将为食用植物油真伪鉴别提供直接的理论和技术支撑，为其它食品真伪鉴别提供良好的理论支撑和借鉴。