Prediction of lipophilicity of polyacenes using quantitative structure-activity relationships

Predictive models for the lipophilicity (logP) of first 25 derivatives of polyacenes are reported. The models are derived from distance-based numerical descriptors which encode information about topology of each compounds in the data set. A new PI-type index called Sadhna index and abbreviated as Sd is introduced for the first time, and its relative correlation potential is established using the results obtained from Wiener (W), Szeged (Sz), first-order Randic connectivity (chi), and Padmakar-Ivan indices. The data show that lipophilicity (logP) is best modelled in bi-parametric model containing PI and Sd indices. The effect due to size, shape, branching, steric and polarity effects on the exhibition of lipophilicity is critically discussed. The predictive ability of the models is discussed on the basis of cross-validation parameters.