Discovery of structure-property relations for molecules via hypothesis-driven active learning over the chemical space

Abstract

Discovery of the molecular candidates for applications in drug targets, biomolecular systems, catalysts, photovoltaics, organic electronics, and batteries, necessitates development of machine learning algorithms capable of rapid exploration of the chemical spaces targeting the desired functionalities. Here we introduce a novel approach for the active learning over the chemical spaces based on hypothesis learning. We construct the hypotheses on the possible relationships between structures and functionalities of interest based on a small subset of data and introduce them as (probabilistic) mean functions for the Gaussian process. This approach combines the elements from the symbolic regression methods such as SISSO and active learning into a single framework. The primary focus of constructing this framework is to approximate physical laws in an active learning regime toward a more robust predictive performance, as traditional evaluation on hold-out sets in machine learning doesn't account for out-of-distribution effects and may lead to a complete failure on unseen chemical space. Here, we demonstrate it for the QM9 dataset, but it can be applied more broadly to datasets from both domains of molecular and solid-state materials sciences.

References in corpus (26)

• Automatic chemical design using a data-driven continuous representation of molecules
• Learning to Plan Chemical Syntheses
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Deep Reinforcement Learning for De-Novo Drug Design
• PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
• Machine learning for molecular simulation
• Less is more: sampling chemical space with active learning
• SISSO: a compressed-sensing method for identifying the best low-dimensional descriptor in an immensity of offered candidates
• Machine Learning Unifies the Modelling of Materials and Molecules
• Atomistic Line Graph Neural Network for Improved Materials Property Predictions
• Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields
• Machine learning prediction errors better than DFT accuracy
• SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
• Gaussian Process Kernels for Pattern Discovery and Extrapolation
• Kernel Interpolation for Scalable Structured Gaussian Processes (KISS-GP)
• Accurate molecular polarizabilities with coupled-cluster theory and machine learning
• Gaussian Processes and Kernel Methods: A Review on Connections and Equivalences
• QM7-X: A comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules
• Machine learning based energy-free structure predictions of molecules (closed and open-shell), transition states, and solids
• Quantum-Mechanical Relation between Atomic Dipole Polarizability and the van der Waals Radius
• Hypothesis Learning in Automated Experiment: Application to Combinatorial Materials Libraries
• Learning molecular energies using localized graph kernels
• Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures
• Molecular Generation with Recurrent Neural Networks (RNNs)
• Heterogeneous Molecular Graph Neural Networks for Predicting Molecule Properties