Generating valid Euclidean distance matrices

Abstract

Generating point clouds, e.g., molecular structures, in arbitrary rotations, translations, and enumerations remains a challenging task. Meanwhile, neural networks utilizing symmetry invariant layers have been shown to be able to optimize their training objective in a data-efficient way. In this spirit, we present an architecture which allows to produce valid Euclidean distance matrices, which by construction are already invariant under rotation and translation of the described object. Motivated by the goal to generate molecular structures in Cartesian space, we use this architecture to construct a Wasserstein GAN utilizing a permutation invariant critic network. This makes it possible to generate molecular structures in a one-shot fashion by producing Euclidean distance matrices which have a three-dimensional embedding.

9 pages, 5 figures

References in corpus (14)

• Semi-Supervised Classification with Graph Convolutional Networks
• Neural Message Passing for Quantum Chemistry
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• SchNet - a deep learning architecture for molecules and materials
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Molecular Graph Convolutions: Moving Beyond Fingerprints
• Progressive Growing of GANs for Improved Quality, Stability, and Variation
• Objective-Reinforced Generative Adversarial Networks (ORGAN) for Sequence Generation Models
• Neural networks-based variationally enhanced sampling
• Riemannian SVRG: Fast Stochastic Optimization on Riemannian Manifolds
• Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules
• PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows
• Generating equilibrium molecules with deep neural networks
• Machine Learning of Energetic Material Properties

Cited by in corpus (9)

• Machine Learning Force Fields
• Inverse design of 3d molecular structures with conditional generative neural networks
• Learning Neural Generative Dynamics for Molecular Conformation Generation
• 3DMolNet: A Generative Network for Molecular Structures
• An End-to-End Framework for Molecular Conformation Generation via Bilevel Programming
• Learning a Continuous Representation of 3D Molecular Structures with Deep Generative Models
• Graph-convolution neural network-based flexible docking utilizing coarse-grained distance matrix
• Auto-Encoding Molecular Conformations
• MOLUCINATE: A Generative Model for Molecules in 3D Space