蚌兰花挥发油化学成分色谱保留値的构效关系研究

基于著名的Kier 连接性指数(mXtV)及电性距离矢量(Md)，使用最佳变量集回归(LBR)建立蚌兰花中20 种有机成分的气相色谱保留时间(tR)的四元数学模型，其传统相关系数(R2)为0.963，逐一剔除法(LOO)交叉验证系数(Q)为0.832。该模型具有高度稳健性与良好的预测能力，影响有机物气相色谱保留时间的主要因素是分子大小与空间形状。结果表明，Kier 逆指数对有机物气相色谱保留时间的表征是合理的、有效的，所建模型能较好解释其递变规律。

Quantitative Structure-retention Relationship Studies of Chemical Constituents in Volatile Oil from Rhoeo spathacea (Sw.) Stearn Flowers

On the basis of Kier s molecular connectivity index (mXtV) and electronegativity distance vector (Md), the retention time of 20 organic components in volatile oil from Rhoeo spathacea (Sw.) Stearn flowers was determined and a four-variable model (QSRR) of gas chromatographic retention time (tR/min) for 20 chemical components was established from mXtV, Md by leaps-andbounds regression (LBR). Traditional correlation coefficient (R) and cross-validation correlation coefficient (Q) of leave-one-out (LOO) were 0.963 and 0.832, respectively. The model was highly reliable and had favorable prediction capability. The dominant factors for retention time were molecular size and spatial shape of organic molecules. The Kier s converse indices and electronegativity distance vector exhibited a good rationality and efficiency in characterizing retention time of organic compounds. This model could elucidate the change trend of retention indices for organic components.