文摘
Accurate in silico models for predicting aqueous solubility are needed in drug design and discovery andmany other areas of chemical research. We present a statistical modeling of aqueous solubility based onmeasured data, using a Gaussian Process nonlinear regression model (GPsol). We compare our results withthose of 14 scientific studies and 6 commercial tools. This shows that the developed model achieves muchhigher accuracy than available commercial tools for the prediction of solubility of electrolytes. On top ofthe high accuracy, the proposed machine learning model also provides error bars for each individual prediction.