Obtaining the desired effect of drugs is highly dependent on their molecular geometries. Thus, the current prevailing paradigm focuses on 3D point-cloud atom representations, utilizing graph neural network (GNN) parametrizations, with rotational symmetries baked in via E(3) invariant layers. We prove that such models must necessarily disregard chirality, a geometric property of the molecules that cannot be superimposed on their mirror image by rotation and translation. Chirality plays a key role in determining drug safety and potency. To address this glaring issue, we introduce a novel field-based representation, proposing reference rotations that replace rotational symmetry constraints. The proposed model captures all molecular geometries including chirality, while still achieving highly competitive performance with E(3)-based methods across standard benchmarking metrics.
@inproceedings{
dumitrescu2025eequivariant,
title={E(3)-equivariant models cannot learn chirality: Field-based molecular generation},
author={Alexandru Dumitrescu and Dani Korpela and Markus Heinonen and Yogesh Verma and Valerii Iakovlev and Vikas Garg and Harri L{\"a}hdesm{\"a}ki},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025},
url={https://openreview.net/forum?id=mXHTifc1Fn}
}