Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models

2020/10/16 by Matthew Ragoza, Ragoza, Matthew, Tomohide Masuda +3 · 7 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · Materials Science · #Biomolecules (q-bio.BM) #Computational Drug Discovery Methods #FOS: Biological sciences #FOS: Computer and information sciences #Graph Theory and Algorithms #Machine Learning (cs.LG) #Machine Learning in Materials Science #Protein Structure and Dynamics #Quantitative Methods (q-bio.QM) #cs.LG #q-bio.BM #q-bio.QM

paper · pdf · doi:10.48550/arxiv.2010.08687

Camera-ready submission to NeurIPS 2020 MLSB workshop

openalex publication_date 2020/10/16 · arxiv created 2020/11/15 · arxiv updated 2020/11/17 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Machine learning in drug discovery has been focused on virtual screening of molecular libraries using discriminative models. Generative models are an entirely different approach that learn to represent and optimize molecules in a continuous latent space. These methods have been increasingly successful at generating two dimensional molecules as SMILES strings and molecular graphs. In this work, we describe deep generative models of three dimensional molecular structures using atomic density grids and a novel fitting algorithm for converting continuous grids to discrete molecular structures. Our models jointly represent drug-like molecules and their conformations in a latent space that can be explored through interpolation. We are also able to sample diverse sets of molecules based on a given input compound and increase the probability of creating valid, drug-like molecules.

Citations

• Generative Adversarial Networks
• ImageNet classification with deep convolutional neural networks
• AtomNet: A Deep Convolutional Neural Network for Bioactivity Prediction in Structure-based Drug Discovery
• Molecular Graph Convolutions: Moving Beyond Fingerprints
• Objective-Reinforced Generative Adversarial Networks (ORGAN) for Sequence Generation Models
• Junction Tree Variational Autoencoder for Molecular Graph Generation

Cited by

• Pharmacophore-based design by learning on voxel grids
• Predictive Feature Caching for Training-free Acceleration of Molecular Geometry Generation
• FlowMol3: Flow Matching for 3D De Novo Small-Molecule Generation
• A 3D Generative Model for Structure-Based Drug Design
• Do we need equivariant models for molecule generation?
• Sampling 3D Molecular Conformers with Diffusion Transformers
• Generating 3D Molecular Structures Conditional on a Receptor Binding Site with Deep Generative Models

Related

• Generating 3D Molecular Structures Conditional on a Receptor Binding Site with Deep Generative Models
• Generating 3D Molecules Conditional on Receptor Binding Sites with Deep Generative Models
• A 3D Generative Model for Structure-Based Drug Design
• Equivariant Shape-Conditioned Generation of 3D Molecules for Ligand-Based Drug Design
• Structure-aware generation of drug-like molecules