Improving Molecular Modeling with Geometric GNNs: An Empirical Study
Abstract
Rapid advancements in machine learning (ML) are transforming materials science by significantly speeding up material property calculations. However, the proliferation of ML approaches has made it challenging for scientists to keep up with the most promising techniques. This paper presents an empirical study on Geometric Graph Neural Networks for 3D atomic systems, focusing on the impact of different (1) canonicalization methods, (2) graph creation strategies, and (3) auxiliary tasks, on performance, scalability and symmetry enforcement. Our findings and insights aim to guide researchers in selecting optimal modeling components for molecular modeling tasks.
Cite
Text
Ramlaoui et al. "Improving Molecular Modeling with Geometric GNNs: An Empirical Study." ICML 2024 Workshops: ML4LMS, 2024.Markdown
[Ramlaoui et al. "Improving Molecular Modeling with Geometric GNNs: An Empirical Study." ICML 2024 Workshops: ML4LMS, 2024.](https://mlanthology.org/icmlw/2024/ramlaoui2024icmlw-improving/)BibTeX
@inproceedings{ramlaoui2024icmlw-improving,
title = {{Improving Molecular Modeling with Geometric GNNs: An Empirical Study}},
author = {Ramlaoui, Ali and Saulus, Théo and Terver, Basile and Schmidt, Victor and Rolnick, David and Malliaros, Fragkiskos D. and Duval, Alexandre AGM},
booktitle = {ICML 2024 Workshops: ML4LMS},
year = {2024},
url = {https://mlanthology.org/icmlw/2024/ramlaoui2024icmlw-improving/}
}