QSPR molecular approach for estimating Henry's law constants of pure compounds in water at ambient conditions

International audienceIn this article, we present a comprehensive quantitative structure-property relationship (QSPR) to estimate the Henry's law constant (H) of pure compounds in water at ambient conditions. This relationship is a multilinear equation containing eight chemical-structure-based parameters. The parameters were selected by the genetic algorithm multivariate linear regression (GA-MLR) method using more than 3000 molecular descriptors. The squared correlation coefficient of the model (R 2) over 1954 pure compounds is equal to 0.983 (logarithmic-based data). Therefore, the model is comprehensive and accurate enough to be used to predict the Henry's law constants of various compounds in water