Informing geometric deep learning with electronic interactions to accelerate quantum chemistry

• Published in: Proceedings of the National Academy of Sciences - PNAS Vol. 119; no. 31; p. e2205221119
• Main Authors: Qiao, Zhuoran, Christensen, Anders S, Welborn, Matthew, Manby, Frederick R, Anandkumar, Anima, Miller, 3rd, Thomas F
• Format: Journal Article
• Language: English
• Published: United States National Academy of Sciences 02-08-2022
• Subjects: Benchmarks; Catalysts; Charge transfer; Chemical properties; Chemical reactions; Deep Learning; Density functional theory; Electronic structure; Electronics; Learning algorithms; Machine Learning; Materials science; Molecular structure; Neural networks; Neural Networks, Computer; Physical Sciences; Quantum chemistry; Small Molecule Libraries; Training; quantum chemistry; machine learning; equivariance
• ISSN: 0027-8424; 1091-6490; 1091-6490
• DOI: 10.1073/pnas.2205221119

Predicting electronic energies, densities, and related chemical properties can facilitate the discovery of novel catalysts, medicines, and battery materials. However, existing machine learning techniques are challenged by the scarcity of training data when exploring unknown chemical spaces. We overcome this barrier by systematically incorporating knowledge of molecular electronic structure into deep learning. By developing a physics-inspired equivariant neural network, we introduce a method to learn molecular representations based on the electronic interactions among atomic orbitals. Our method, OrbNet-Equi, leverages efficient tight-binding simulations and learned mappings to recover high-fidelity physical quantities. OrbNet-Equi accurately models a wide spectrum of target properties while being several orders of magnitude faster than density functional theory. Despite only using training samples collected from readily available small-molecule libraries, OrbNet-Equi outperforms traditional semiempirical and machine learning-based methods on comprehensive downstream benchmarks that encompass diverse main-group chemical processes. Our method also describes interactions in challenging charge-transfer complexes and open-shell systems. We anticipate that the strategy presented here will help to expand opportunities for studies in chemistry and materials science, where the acquisition of experimental or reference training data is costly.