【摘要】 在硅胶薄层板上用氯仿 -乙酸乙酯 -甲醇 -水 (体积比为 15∶ 4 0∶ 2 2∶ 10 )作展开剂 ,测定了常见人参皂甙的Rf值。为了研究它们的结构与保留值之间的关系 ,对它们的 17种结构参数进行了计算。除了拓扑指数和理化参数外 ,引入了代表构成人参皂甙的母体化合物种类并反映它们分子极性特征的新参数“E”。通过相关分析优化选出范德华分子表面积 AW、拓扑指数0 B和参数 E,建立了多参数线性回归方程 ,较好地描述了在正相薄层色谱中常见人参皂甙结构与保留值之间的关系 ,并与人工神经网络方法进行了比较。结果表明 ,非线性方法更适合于该研究 ,为进一步采用人工神经网络方法研究正相薄层色谱中人参皂甙结构与保留之间的关系打下了良好基础更多还原

【Abstract】 The TLC retention of ten ginsenosides has been determined on silica gel layers developed with chloroform ethyl acetate methanol water(15∶40∶22∶10, V/V ) as the mobile phase. For the quantitative structure\|retention relationship (QSRR) study seventeen structure parameters of the ten ginsenosides were calculated, including topologic indices and physicochemical properties. Particularly, a novel parameter “ E ” was presented according to the differences of the parent substances of which the ginsenosides are composed. Of these parameters, A\-w, 0B and E were selected by optimization and employed in the derived multi parameter regression equation. The relationships between the retention and the molecular structures have been discussed.更多还原