基于多源化学信息融合技术的云茯苓资源评价研究

Edible and medicinal mushrooms get more and more popular and become a research focus because of the edible and medicinal effects. Quality control and safety of the resources of these mushrooms is an important premise to guarantee the quality of mushroom products. So, it is necessary to establish a comprehensive and systematic technique in order to assess the quality of these resources. Multi-source chemical information fusion could contain a large number of information on the chemical components and realize the diversification of analysis and detection methods. Furthermore, it could also remedy the limitation of a simple evaluation pattern and acquire the intuitionistic, systematic, comprehensive and accurate evaluating effectiveness. Based on collecting a total of 80 populations, 780 Wolfiporia cocos samples from Yunnan Province, this project intends to obtain the multi-source chemical information of different parts of wild and cultured W. cocos samples which obtained from different regions by using UV, FTIR and UPLC methods and determine the contents of characteristic components, simultaneously. Combined with chemometrics, this project aims to (1) establish the qualitative evaluation system of W. cocos based on multi-source chemical information fusion, (2) screen the characteristic constituents which have close relationships with the quality of W. cocos and investigate the accumulation and variation of these components and (3) set up the information model of characteristic components to perform chemical pattern recognition. Then the quality evaluation system which may be suitable for the international requirements will be established. The results could provide the scientific basis for improving the quality standard of W. cocos. Meanwhile, it could also provide the reasonable foundation for development and utilization on the resources of edible and medicinal mushrooms.

食（药）用真菌因同时兼具食用和药用价值备受关注，成为研究开发的热点。其质量安全可控有利于后期产品质量从源头上得到保障，有必要建立全面、系统的方法对食（药）用真菌资源进行质量评价研究。多源化学信息融合技术能涵盖大量化学成分信息，实现分析方法和检测方式多元化，弥补单一评价模式的缺陷，达到系统全面的评价效果。本项目拟在收集云南地区80个居群780份茯苓样品基础上，采用紫外光谱、红外光谱和超高效液相色谱法获得不同产区、不同部位、野生与栽培云茯苓化学信息，同时测定云茯苓特征性成分含量，结合化学计量学方法：（1）构建云茯苓多源化学信息融合定性评价体系；（2）筛选表征云茯苓品质的特征性成分，探明特征性成分的累积与变化规律；（3）构建云茯苓特征性成分信息模型，建立与国际接轨的质量评价体系，为完善茯苓质量标准提供科学依据，同时为食（药）用真菌资源合理开发利用提供理论基础。

化学成分能够反映云茯苓内在属性，是其品质形成及发挥治疗效果的物质基础。药用真菌化学组成成分复杂，易受产地、生长模式等多种因素影响，如何选择合理的表征方式对化学成分进行解析，是药用资源化学评价研究亟待解决的关键问题之一。本项目以云茯苓为研究对象，采用多源化学信息融合技术分析，研究发现：不同产地、部位之间化学组成和茯苓酸含量具有差异，茯苓不同部位化学成分和含量的差异大于采自不同产地相同部位的样品；产地来源对茯苓皮中化学成分的影响大于白茯苓；同一产地的样品，其质量亦存在差异；栽培茯苓皮中化学成分可能与其生长环境的土壤有关，产地土壤颜色相同的样品，质量差异较小；白茯苓的质量比茯苓皮有更好的一致性；栽培样品表皮和内部的去氢土莫酸、茯苓新酸A和松苓新酸含量与野生样品的显著不同；ATR-FTIR结合PLSR对茯苓中的茯苓新酸A和松苓新酸含量能够成功预测，在特征提取中VIP优于间隔PLS，紫外光谱和中红外光谱分别是快速预测茯苓中茯苓新酸A、去氢土莫酸和松苓新酸含量和产地鉴别的有效手段；同步SD 2DCOS光谱图像更适用于白茯苓和茯苓皮的特征波段小且复杂混合系统的识别和分析。从有效性和可测性角度整合分析显示，松苓新酸、茯苓新酸A和茯苓酸可作为白茯苓和茯苓皮的初步潜在质量标志物；茯苓皮中三种三萜酸的含量均高于茯苓核，其中野生茯苓皮的含量最高。茯苓的适生区主要分布在中国云南省，约占中国适生区总面积的49%。此外，建立了傅立叶变换红外光谱结合偏最小二乘判别分析（PLS-DA）和残差神经网络（ResNet）技术对云茯苓不同产地和部位鉴别的新方法。通过项目的实施，在国内外刊物发表研究论文16篇，其中SCI收录15篇。培养硕士研究生2人，获硕士学位1人；入选云南省中青年学术和技术带头人后备人才1人、云南省高层次人才培养支持计划“青年拔尖人才”专项1人；获云南省青年科技奖1人。