Modeling quantitative structure-property relationships to predict the environmental properties of some iodinated organic compounds

The purpose of this research is modeling and predicting the physicochemical and environmental properties of iodinated organic compounds theoretically and help to design and manufacture new compounds. In this article, an attempt has been made to model and predict the physicochemical properties of 43 iodinated organic compounds using multiple linear regression methods. In this study, more than 255 different topological indices were used and all of them reported in a data matrix. Multiple linear regression techniques were used to obtain a new quantitative structureproperty relationship model with high statistical parameters to predict the properties of octanolwater partition coefficient, aqueous solubility, air solubility, vaporization enthalpy, solution enthalpy and normal boiling point of 43 organic iodinated compounds and then a genetic algorithm was applied to select suitable indices. This study shows that the genetic algorithm is very useful in quantitative structure-property relationship analysis and provides a new approach to building better models. The cross-sectional evaluation method of leaving an exit used in this study ensured that the implemented models have acceptable stability and validity. Also, the prediction ability of the model was confirmed by the external evaluation method.