Evidential Deep Learning for Guided Molecular Property Prediction and Discovery
Name
acscentsci.1c00546.pdf
Description
Published version
Size
2.62 MB
Format
Adobe PDF
Checksum (MD5)
68176c4240f03b06693d08ed3f9d7305
Author(s)
Soleimany, Ava P
Amini, Alexander
Goldman, Samuel
Rus, Daniela
Bhatia, Sangeeta N
Coley, Connor W
Date Issued
2021
Journal
ACS Central Science
Publisher
American Chemical Society (ACS)
Citation
Soleimany, Ava P, Amini, Alexander, Goldman, Samuel, Rus, Daniela, Bhatia, Sangeeta N et al. 2021. "Evidential Deep Learning for Guided Molecular Property Prediction and Discovery." ACS Central Science, 7 (8).
Version
Final published version
Abstract
While neural networks achieve state-of-the-art performance for many molecular modeling and structure-property prediction tasks, these models can struggle with generalization to out-of-domain examples, exhibit poor sample efficiency, and produce uncalibrated predictions. In this paper, we leverage advances in evidential deep learning to demonstrate a new approach to uncertainty quantification for neural network-based molecular structure-property prediction at no additional computational cost. We develop both evidential 2D message passing neural networks and evidential 3D atomistic neural networks and apply these networks across a range of different tasks. We demonstrate that evidential uncertainties enable (1) calibrated predictions where uncertainty correlates with error, (2) sample-efficient training through uncertainty-guided active learning, and (3) improved experimental validation rates in a retrospective virtual screening campaign. Our results suggest that evidential deep learning can provide an efficient means of uncertainty quantification useful for molecular property prediction, discovery, and design tasks in the chemical and physical sciences.
MIT Department
Massachusetts Institute of Technology. Department of Chemical Engineering
Harvard University--MIT Division of Health Sciences and Technology
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Massachusetts Institute of Technology. Computational and Systems Biology Program
Howard Hughes Medical Institute
Terms of Use
Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International Licens
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1021/ACSCENTSCI.1C00546
