GeoRecon: Graph-Level Representation Learning for 3D Molecules via Reconstruction-Based Pretraining

TL;DR

GeoRecon introduces a graph-level pretraining method for 3D molecules that enhances global structural understanding and improves performance on molecular property prediction benchmarks.

Contribution

It proposes a novel graph-level reconstruction pretraining framework that captures holistic molecular structures beyond local atomic environments.

Findings

Outperforms baseline models on QM9, MD17, MD22, and 3BPA datasets.

Encourages learning of smooth, transferable latent spaces.

Enhances global structural feature encoding in molecular representations.

Abstract

The pretraining-finetuning paradigm has powered major advances in domains such as natural language processing and computer vision, with representative examples including masked language modeling and next-token prediction. In molecular representation learning, however, pretraining tasks remain largely restricted to node-level denoising, which effectively captures local atomic environments but is often insufficient for encoding the global molecular structure critical to graph-level property prediction tasks such as energy estimation and molecular regression. To address this gap, we introduce GeoRecon, a graph-level pretraining framework that shifts the focus from individual atoms to the molecule as an integrated whole. GeoRecon formulates a graph-level reconstruction task: during pretraining, the model is trained to produce an informative graph representation that guides geometry reconstruction while inducing smoother and more transferable latent spaces. This encourages the learning of coherent, global structural features beyond isolated atomic details. Without relying on external supervision, GeoRecon generally improves over backbone baselines on multiple molecular benchmarks including QM9, MD17, MD22, and 3BPA, demonstrating the effectiveness of graph-level reconstruction for holistic and geometry-aware molecular embeddings.