3D matters! 3D-RISM and 3D convolutional neural network for accurate bioaccumulation prediction

Sergey Sosnin, Maksim Misin, David S. Palmer, Maxim V. Fedorov

Research output: Contribution to journalArticle

7 Citations (Scopus)
5 Downloads (Pure)


In this work, we present a new method for predicting complex physicalchemical properties 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 threedimensional reference interaction site model (3D-RISM). We have shown that the method allows one to achieve a good accuracy of prediction of bioconcentration factor (BCF) which is difficult to predict by direct application of methods of molecular theory or simulations. 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 theory-based models.
Original languageEnglish
Article number32LT03
Number of pages7
JournalJournal of Physics: Condensed Matter
Publication statusPublished - 19 Jul 2018



  • deep learning (DL)
  • convolution neural network (CNN)
  • 3DRISM
  • machine learning
  • bioaccumulation
  • molecular solvation
  • computational chemistry
  • molecular simulation

Cite this