发酵玉米芯多糖、还原糖含量近红外模型的建立  

Establishment of near infrared model for the content of polysaccharide and reducing sugar in fermented corncob

作　　者：张燕[1,2] 王园 杜涓 李施垚 郑越 王春媛 齐景伟 安晓萍[1,2] ZHANG Yan;WANG Yuan;DU Juan;LI Shiyao;ZHENG Yue;WANG Chunyuan;QI Jingwei;AN Xiaoping(College of Animal Science,Inner Mongolia Agricultural University,Hohhot,Inner Mongolia 010018,China;Inner Mongolia Herbivorous Livestock Feed Engineering and Technology Research Center,Hohhot,Inner Mongolia 010018,China)

机构地区：[1]内蒙古农业大学动物科学学院,内蒙古呼和浩特010018 [2]内蒙古自治区草食家畜饲料工程技术研究中心,内蒙古呼和浩特010018

出　　处：《中国饲料》2024年第7期101-108,共8页China Feed

基　　金：内蒙古自治区关键技术攻关计划项目(2020GG0030);内蒙古自治区科技重大专项(2021ZD0024-4)。

摘　　要：本实验旨在应用近红外光谱技术(NIRS)结合化学计量学法快速预测发酵玉米芯中多糖、还原糖含量,为定量检测发酵玉米芯多糖、还原糖含量提供理论依据以及为玉米芯深加工利用提供技术支持。以105份微生物发酵的玉米芯为供试材料,采用苯酚-硫酸法和DNS法分别测定多糖和还原糖含量。利用偏最小二乘法(PLS),通过不同预处方式和不同波长建立发酵玉米芯多糖、还原糖含量的近红外分析模型。结果表明:多糖模型采用标准正态变量变换(SNV)+1阶导数的方法对全谱图进行预处理的效果较好,优化后的模型决定系数(R2)、校正均方根误差(RMSEC)、校准标准差(SEC)分别为0.82、9.28、9.34,其相对分析误差(PRD)为2.37;还原糖模型采用1阶导数+标准正态变量变换(SNV)+去趋势化(Detrend)的方法对全谱图进行预处理的效果较好,优化后的模型决定系数(R2)、校正均方根误差(RMSEC)、校准标准差(SEC)分别为0.84、4.03、4.04,其相对分析误差(PRD)为2.48;预测集决定系数分别为0.85、0.88。本研究构建的NIRS模型校正和交互验证决定系数均较大,相对分析误差均大于2,说明模型预测性能较好,建立的模型有助于发酵玉米芯多糖、还原糖含量活性成分的筛选。In order to provide theoretical basis for quantitative determination of polysaccharide and reducing sugar content in fermented corn cob and technical support for deep processing and utilization of corn cob,near infrared spectroscopy combined with stoichiometry was used to rapidly predict the content of polysaccharide and reducing sugar in fermented corn cob.The contents of polysaccharide and reducing sugar were determined by phenol-sulfuric acid method and DNS method.Partial least squares(PLS)method was used to establish the near infrared analysis model of the contents of polysaccharide and reducing sugar in fermented corn cob by different pretreatment methods and different wavelength.The results showed that:the polysaccharide model was pretreated by SNV+1 derivative method.The coefficient of determination(R2),root mean square error(RMSEC)and standard deviation(SEC)of the optimized model were 0.82,9.28 and 9.34,respectively,and the relative analysis error(PRD)was 2.37.The first-order+SNV+Detrend method was used to pretreat the full spectrum of the reducing sugar model.The coefficient of determination(R2),corrected root mean square error(RMSEC)and calibrated standard deviation(SEC)of the optimized model were 0.84,4.03 and 4.04,respectively.The relative analysis error(PRD)was 2.48.The determination coefficients of prediction set were 0.85 and 0.88,respectively.The determination coefficients of the NIRS model established in this study were both large,and the relative analysis errors were both greater than 2,indicating that the model had good prediction performance,and the established model was helpful to the screening of active components of polysaccharide and reducing sugar content in fermented corn cob.

关 键 词：玉米芯 多糖 还原糖 近红外模型 偏最小二乘法 

分 类 号：S816.17[农业科学—饲料科学]