Bosque Ramón, Sales Joaquim, Bosch Elisabeth, Rosés Martí, García-Alvarez-Coque M C, Torres-Lapasió J R
Departament de Química Inorgànica, Universitat de Barcelona, Martí i Franquès, 1, 08028-Barcelona, Spain.
J Chem Inf Comput Sci. 2003 Jul-Aug;43(4):1240-7. doi: 10.1021/ci0340660.
A Quantitative Structure-Property Relationship (QSPR) model is developed to calculate the solute polarity parameter p of a set of 233 compounds of a very different chemical nature. The proposed model, derived from multiple linear regression, contains four descriptors calculated from the molecular structure and the well-known hydrophobicity parameter log P(o/w). According to the statistics obtained with the prediction set, the model has a very good prediction capacity (R(2) = 0.954, F = 889, n = 45, and SD = 0.27). The study shows that log P(o/w) and hydrogen bond acidity of the solutes are the most relevant descriptors to predict p values. This p parameter is embodied in a general equation to predict retention in reversed-phase liquid chromatography (RP-HPLC). It describes analyte retention exclusively on the basis of mobile phase/analyte/stationary phase polar interactions. Equations and procedures to determine polarity of both chromatographic phases had been successfully developed previously. Therefore, the proposed QSPR model for p estimation becomes a very useful tool in RP-HPLC optimization of procedures and methods in the everyday analytical work.
建立了一种定量结构-性质关系（QSPR）模型，用于计算233种化学性质差异很大的化合物的溶质极性参数p。该模型源自多元线性回归，包含四个由分子结构和著名的疏水性参数log P(o/w)计算得出的描述符。根据预测集得到的统计数据，该模型具有很好的预测能力（R(2) = 0.954，F = 889，n = 45，SD = 0.27）。研究表明，溶质的log P(o/w)和氢键酸度是预测p值最相关的描述符。这个p参数体现在一个预测反相液相色谱（RP-HPLC）保留时间的通用方程中。它仅基于流动相/分析物/固定相的极性相互作用来描述分析物的保留。此前已成功开发出确定两种色谱相极性的方程和程序。因此，所提出的用于估计p的QSPR模型成为日常分析工作中RP-HPLC程序和方法优化的非常有用的工具。
