Recent advances in machine learning have enabled generative models for both optimization and de novo generation of drug candidates with desired properties. Previous generative models have focused on producing SMILES strings or 2D molecular graphs, while attempts at producing molecules in 3D have focused on reinforcement learning (RL), distance matrices, and pure atom density grids. Here we present MOLUCINATE (MOLecUlar ConvolutIoNal generATive modEl), a novel architecture that simultaneously generates topological and 3D atom position information. We demonstrate the utility of this method by using it to optimize molecules for desired radius of gyration. In the future, this model can be used for more useful optimization such as binding affinity against a protein target.
Michael Brocidiacono (Unaffiliated)
David Koes (University of Pittsburgh)
I develop novel computational algorithms and build full-scale systems to support rapid and inexpensive drug discovery while simultaneously applying these methods to develop novel therapeutics. I seek to unlock the power of computation and machine learning to solve challenging, real world problems and am a staunch advocate of open source software and open science.
More from the Same Authors
2021 : Exploring ∆∆G prediction with Siamese Networks »
Andrew McNutt · David Koes
2020 : Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models »
Matthew Ragoza · Tomohide Masuda · David Koes
2020 : Contributed Talk - Generating 3D Molecular Structures Conditional on a Receptor Binding Site with Deep Generative Models »
Tomohide Masuda · Matthew Ragoza · David Koes
2020 : Contributed Talk - SidechainNet: An All-Atom Protein Structure Dataset for Machine Learning »
Jonathan King · David Koes
2020 : David Ryan Koes »