Multi-Level Fusion Graph Neural Network for Molecule Property Prediction

TL;DR

This paper introduces MLFGNN, a novel graph neural network that combines attention mechanisms and molecular fingerprints to better capture local and global molecular structures, significantly improving property prediction accuracy.

Contribution

The work presents a multi-level fusion approach integrating GATs, Graph Transformers, and molecular fingerprints, with an adaptive attention mechanism for enhanced molecular representation.

Findings

Outperforms state-of-the-art methods on benchmark datasets

Effectively captures chemical patterns relevant to tasks

Improves both classification and regression accuracy

Abstract

Accurate prediction of molecular properties is essential in drug discovery and related fields. However, existing graph neural networks (GNNs) often struggle to simultaneously capture both local and global molecular structures. In this work, we propose a Multi-Level Fusion Graph Neural Network (MLFGNN) that integrates Graph Attention Networks and a novel Graph Transformer to jointly model local and global dependencies. In addition, we incorporate molecular fingerprints as a complementary modality and introduce a mechanism of interaction between attention to adaptively fuse information across representations. Extensive experiments on multiple benchmark datasets demonstrate that MLFGNN consistently outperforms state-of-the-art methods in both classification and regression tasks. Interpretability analysis further reveals that the model effectively captures task-relevant chemical patterns, supporting the usefulness of multi-level and multi-modal fusion in molecular representation learning.