Predicting bioaccumulation using molecular theory: A machine learning approach (1710.08174v1)

Abstract: In this work, we present a new method for predicting bioaccumulation factor of organic molecules. The approach utilizes 3D convolutional neural network (ActivNet4) that uses solvent spatial distributions around solutes as input. These spatial distributions are obtained by a molecular theory called three-dimensional reference interaction site model (3D-RISM). We have shown that the method allows one to achieve a good accuracy of prediction. Our research demonstrates that combination of molecular theories with modern machine learning approaches can be effectively used for predicting properties that are otherwise inaccessible to purely physics-based models.