Chromatographic Retention Times of Polychlorinated Biphenyls: from Structural Information to Property Characterization

The paper presents a unitary approach of the use of a Molecular DescriptorsFamily in structure-property/activity relationships, particularly in modelling thechromatographic retention times of polychlorinated biphenyls. Starting from molecularstructure, viewed as a graph, and considering the bonds and bond types, atom types andoften the 3D geometry of the molecule, a huge family of molecular descriptors called MDFwas calculated. A preliminary selection of MDF members was done by simple linearregression (LR) against the measured property. The best fitted MDF subset is thensubmitted to multivariate linear regression (MLR) analysis in order to find the best pairs ofMDF members that produce a reliable QSPR (Quantitative Structure-PropertyRelationship) model. The predictive capability was finally tested by randomly splitting ofdata into training and test sets. The best obtained models are presented and the results arediscussed