血橙果实香气成分色谱保留指数预测

Prediction of retention indices of some components from Blood Orange Fruit

摘要: 应用分子电性距离矢量（MEDV）对血橙中的45种挥发性组分的气相色谱保留指数（RI）进行了定量结构-色谱保留关系（QSRR）研究。通过多元线性回归（MLR）和逐步回归（SMR）得到了该类化合物气相色谱保留指数（RI）的预测模型,模型的复相关系数（R）为0.947,标准误差（SD）为105.039。留一法（Leave-one-out,LOO）交互检验对模型进行了评价,得到的复相关系数（RCV）为0.923,标准误差（SDCV）为126.394。结果表明所建模型具有良好的预测能力和稳定性。

Abstract: Quantitative structure- retention relationship ( QSRR) study of the retention indices of 45 components of Blood Orange Fruit was performed by using called the molecular electronegativity- distance vector ( MEDV) .Through multiple linear regression ( MLR) and stepwise regression ( SMR) , a prediction model of gas chromatography retention indices ( RI) of the compounds in was obtained.The multiple correlation coefficient ( R) of the model was 0.947, and the standard deviation ( SD) was 105.039.Then the model was evaluated by performing the cross validation with the leave- one- out ( LOO) procedure and the cross validation multiple correlation coefficient ( RCV) was 0.923, and the standard deviation ( SDCV) was 126.394.The results showed that the model had good predictive ability and stability.