AUGUR, A flexible and efficient optimization algorithm for identification of optimal adsorption sites

TL;DR

The paper introduces AUGUR, a novel optimization pipeline combining graph neural networks and Gaussian processes to efficiently identify optimal adsorption sites with uncertainty quantification, applicable to molecules of varying sizes.

Contribution

AUGUR is a flexible, symmetry-aware optimization method that requires fewer iterations and can be applied universally without hand-crafted features.

Findings

Determines optimal adsorption sites with fewer iterations than existing methods.

Handles molecules of different sizes using the same model.

Provides uncertainty quantification in predictions.

Abstract

In this paper, we propose a novel flexible optimization pipeline for determining the optimal adsorption sites, named AUGUR (Aware of Uncertainty Graph Unit Regression). Our model combines graph neural networks and Gaussian processes to create a flexible, efficient, symmetry-aware, translation, and rotation-invariant predictor with inbuilt uncertainty quantification. This predictor is then used as a surrogate for a data-efficient Bayesian Optimization scheme to determine the optimal adsorption positions. This pipeline determines the optimal position of large and complicated clusters with far fewer iterations than current state-of-the-art approaches. Further, it does not rely on hand-crafted features and can be seamlessly employed on any molecule without any alterations. Additionally, the pooling properties of graphs allow for the processing of molecules of different sizes by the same model. This allows the energy prediction of computationally demanding systems by a model trained on comparatively smaller and less expensive ones