Enhancing Molecular Property Predictions by Learning from Bond Modelling and Interactions

2026-02-28

AI Summary

Traditional atom-centered models simplify chemical bonds as pairwise interactions, failing to capture complex bond-order effects such as resonance and stereoselectivity, thereby limiting the accuracy of molecular property prediction. To address this, this work proposes DeMol, a dual-graph framework that introduces a bond-centered perspective for the first time. By concurrently modeling atomic and bond graphs, DeMol employs a multi-scale Double-Helix module to jointly learn high-order interactions among atoms, bonds, and their combinations. Leveraging information-theoretic analysis, we quantify the information gain from bond-centered representations and enhance geometric consistency through covalent radius regularization. The method achieves new state-of-the-art results across multiple benchmarks—including PCQM4Mv2, OC20 IS2RE, QM9, and MoleculeNet—demonstrating significantly improved predictive performance.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Knowledge Representation and Reasoning: Computational Complexity of Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Molecule representation learning is crucial for understanding and predicting molecular properties. However, conventional atom-centric models, which treat chemical bonds merely as pairwise interactions, often overlook complex bond-level phenomena like resonance and stereoselectivity. This oversight limits their predictive accuracy for nuanced chemical behaviors. To address this limitation, we introduce \textbf{DeMol}, a dual-graph framework whose architecture is motivated by a rigorous information-theoretic analysis demonstrating the information gain from a bond-centric perspective. DeMol explicitly models molecules through parallel atom-centric and bond-centric channels. These are synergistically fused by multi-scale Double-Helix Blocks designed to learn intricate atom-atom, atom-bond, and bond-bond interactions. The framework's geometric consistency is further enhanced by a regularization term based on covalent radii to enforce chemically plausible structures. Comprehensive evaluations on diverse benchmarks, including PCQM4Mv2, OC20 IS2RE, QM9, and MoleculeNet, show that DeMol establishes a new state-of-the-art, outperforming existing methods. These results confirm the superiority of explicitly modelling bond information and interactions, paving the way for more robust and accurate molecular machine learning.

Problem

• molecular property prediction
• bond modelling
• atom-centric models
• resonance
• stereoselectivity

Innovation

• bond-centric modeling
• dual-graph framework
• Double-Helix Blocks
• molecular representation learning
• geometric regularization