Retention data from normal-phase thin-layer chromatography in characterization of some 1,6-anhydrohexose and D-aldopentose derivatives by QSRR method

Abstract

© 2015 Taylor & Francis Group, LLC. The relationship between retention behavior of eleven 1,6-anhydrohexose and D-aldopentose derivatives and their molecular characteristics was studied using quantitative structure-retention relationships (QSRR) approach. Retention parameters RM⁰, obtained by normal-phase thin-layer chromatography, were correlated with molecular and in silico absorption, distribution, metabolism and excretion (ADME) descriptors. For describing the retention behavior of investigated molecules and determining the similarities between molecules, principal component analysis (PCA) was performed, followed by hierarchical cluster analysis (HCA) and multiple linear regression (MLR). For both sets of descriptors, PCA resulted in a model with the two significant principal components (PCs). HCA was conducted to confirm the grouping of the compounds already obtained by PCA. MLR equations were established for both sets of descriptors and statistical quality of the generated models was determined by standard statistical measures and cross-validation parameters. According to statistical validation, two very good models with molecular and one with in silico ADME descriptors were obtained. Very good predictive ability of the established mathematical models allows us to estimate retention behavior of structurally similar compounds and to understand their behavior in similar chromatographic systems.