脂肪醇气相色谱保留指数的预测与估算

通过对46个脂肪醇在不同固定相不同柱温下的631个样本的气相色谱保留指数值(RI)与其部分参数:拓扑指数(mQ)、定位基参数(Sox)、固定液极性值(CP)及柱温(T)建立定量结构-色谱保留相关(QSRR)模型。分别利用多元线性回归(MLR)、偏最小二乘回归(PLSR)、人工神经网络(ANN)建模,同时采用内部及外部双重验证的办法对所得模型稳定性能进行深入分析和检验,建模计算值、留一法(LOO)交互检验(CV)预测值和外部样本预测值的复相关系数Rmax、QLOO和Rexi分别为0．988,0．987和0．989(MLR);0．988、0．975和0．989(PLSR);0．989、0．989和0．991(ANN)。结果表明:所建定量结构保留相关(QSRR)模型具有良好的稳定性和预测能力,较好地揭示了脂肪醇化合物在不同固定相不同柱温上气相色谱保留指数的变化规律。

Estimation and prediction of the chromatographic retention of fatty alcohols

The article studied the quantitative structure relationship ( QSRR) model of the gas chromatographic retention indices (RI) of 631 data belong to 46 fatty alcohols at different column temperatures on different stationary phases with topological index(mQ) orientating group parameter( Sax) , immobile liquid polarity values(CP) and column temperature( T). Here three models of the gas chromatographic retention were built by multiple linear regression(MLR) , partial least square regression( PLS) and artificial neural network ( ANN) , the estimation stability and generalization ability of these models were strictly analyzed by both internal and external validations. The correlation coefficient( R2) of established MLR, PLS and ANN models, Leave-One-Out(LOO) Cross-Validation(CV) , predicted values versus experimental ones of external samples were 0. 988, 0.987, and 0. 989( MLR) ; 0.988, 0.975 and 0. 989(PLS) ; 0. 989, 0. 989 and 0. 991 ( ANN) , respectively. These models can better elucidate the change rule of the gas chromatographic retention indices for fatty alcohols. The results showed that the quantitative structure relationship (QSRR) model have good stability and predictability.