Physics-Constrained Predictive Molecular Latent Space Discovery with Graph Scattering Variational Autoencoder

2020/09/29 by Navid Shervani-Tabar, Nicholas Zabaras, Shervani-Tabar, Navid +1
Biochemistry, Genetics and Molecular Biology · Computer Science · Materials Science · Mathematics · Physics and Astronomy · #Algorithm #Artificial intelligence #Artificial neural network #Autoencoder #Chemical Physics (physics.chem-ph) #Computational Drug Discovery Methods #Computer science #Decoding methods #FOS: Computer and information sciences #FOS: Physical sciences #Formalism (music) #Generative grammar #Generative model #Graph #Inference #Machine Learning (stat.ML) #Machine Learning in Materials Science #Machine learning #Mathematics #Molecular graph #Protein Structure and Dynamics #Theoretical computer science #Topology (electrical circuits) #physics.chem-ph #stat.ML

paper · pdf · doi:10.48550/arxiv.2009.13878

openalex publication_date 2020/09/29 · openalex created_date 2020/10/08 · arxiv created 2021/02/11 · arxiv updated 2021/02/12 · openalex updated_date 2026/07/28

Abstract

Recent advances in artificial intelligence have propelled the development of innovative computational materials modeling and design techniques. Generative deep learning models have been used for molecular representation, discovery, and design. In this work, we assess the predictive capabilities of a molecular generative model developed based on variational inference and graph theory in the small data regime. Physical constraints that encourage energetically stable molecules are proposed. The encoding network is based on the scattering transform with adaptive spectral filters to allow for better generalization of the model. The decoding network is a one-shot graph generative model that conditions atom types on molecular topology. A Bayesian formalism is considered to capture uncertainties in the predictive estimates of molecular properties. The model's performance is evaluated by generating molecules with desired target properties.

Citations

• NICE: Non-linear Independent Components Estimation
• Tight Wavelet Frames on Multislice Graphs
• The Emerging Field of Signal Processing on Graphs: Extending High-Dimensional Data Analysis to Networks and Other Irregular Domains
• Bootstrap Methods: Another Look at the Jackknife
• Wavelets on graphs via spectral graph theory
• Organic solar cells: An overview
• Understanding Deep Convolutional Networks
• Learning to SMILE(S)
• Semi-Supervised Classification with Graph Convolutional Networks
• Variational Graph Auto-Encoders
• Geometric deep learning: going beyond Euclidean data
• Grammar Variational Autoencoder
• Covariant Compositional Networks For Learning Graphs
• Learning Deep Generative Models of Graphs
• Junction Tree Variational Autoencoder for Molecular Graph Generation
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• N-body Networks: a Covariant Hierarchical Neural Network Architecture for Learning Atomic Potentials
• Deep learning for molecular design - a review of the state of the art
• A Two-Step Graph Convolutional Decoder for Molecule Generation
• Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules
• Stability of Graph Scattering Transforms
• A Deep Learning Approach to Antibiotic Discovery

Related

• NeVAE: A Deep Generative Model for Molecular Graphs
• Conditional Constrained Graph Variational Autoencoders for Molecule Design
• Continuous Representation of Molecules Using Graph Variational Autoencoder
• 3DMolNet: A Generative Network for Molecular Structures
• RGCVAE: Relational Graph Conditioned Variational Autoencoder for Molecule Design