Using pretrained graph neural networks with token mixers as geometric featurizers for conformational dynamics
This work provides a more robust and less manual method for analyzing conformational dynamics in molecular simulations, benefiting researchers working with complex molecular systems.
The paper addresses the challenge of identifying low-dimensional features for molecular simulation dynamics by introducing geom2vec, which uses pretrained graph neural networks (GNNs) as universal geometric featurizers. This approach captures interpretable relationships between structural units and enables analysis of larger molecular graphs without extensive manual tuning or fine-tuning.
Identifying informative low-dimensional features that characterize dynamics in molecular simulations remains a challenge, often requiring extensive manual tuning and system-specific knowledge. Here, we introduce geom2vec, in which pretrained graph neural networks (GNNs) are used as universal geometric featurizers. By pretraining equivariant GNNs on a large dataset of molecular conformations with a self-supervised denoising objective, we obtain transferable structural representations that are useful for learning conformational dynamics without further fine-tuning. We show how the learned GNN representations can capture interpretable relationships between structural units (tokens) by combining them with expressive token mixers. Importantly, decoupling training the GNNs from training for downstream tasks enables analysis of larger molecular graphs (such as small proteins at all-atom resolution) with limited computational resources. In these ways, geom2vec eliminates the need for manual feature selection and increases the robustness of simulation analyses.