arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.03046physics.chem-ph
收藏

机器学习势导出的局部描述符对分子性质预测中三维图神经网络的影响

Impact of Local Descriptors Derived from Machine Learning Potentials in Graph Neural Networks for Molecular Property Prediction

Department of Chemistry and Biochemistry, School of Advanced Science and Engineering, Waseda University(早稻田大学先进理工学研究科化学·生物化学系);Waseda Research Institute for Science and Engineering, Waseda University(早稻田大学理工学研究院);AI Innovation Department, ENEOS Holdings, Inc.(ENEOS控股株式会社AI创新部)

Ryoichi Uchiyama, Yuya Nakajima, Yuta Tanaka, Junji Seino

AI总结:

本研究通过将机器学习势导出的局部描述符整合到3D GNN中,提升分子性质预测的准确性,尤其在QM9和tmQM数据集上表现优异。

摘要:

在本研究中,我们提出了一种框架,旨在通过将从大规模预训练的机器学习势中获得的局部描述符整合到三维图神经网络(3D GNNs)中,以提高分子性质预测的性能。作为示例,我们通过将通过Matlantis获得的首选势(PFP)特征导出的描述符整合到等变图神经网络(EGNN)中,开发了EGNN-PFP模型,并对其有效性进行了评估。当在包含小有机分子的QM9数据集上进行测试时,所提出的模型在12种分子性质中的11种上比原始EGNN模型和没有PFP导出描述符的基线模型表现出更高的准确性。此外,当在包含过渡金属配合物的tmQM数据集上进行评估时,在所有五个目标属性上观察到显著的性能提升,表明过渡金属周围局部原子环境的重要性。本质上,所提出的方法可以适应任何3D GNN架构,并且当与不断演进的GNN架构整合时,预测准确性有望进一步提高。

英文摘要:

In this study, we present a framework aimed at enhancing molecular property prediction through the integration of local descriptors obtained from large-scale pretrained machine learning potentials into three-dimensional graph neural networks (3D GNNs). As an illustration, we developed an EGNN-PFP model by integrating descriptors derived from the preferred potential (PFP) features, acquired through Matlantis, into an equivariant graph neural network (EGNN), and evaluated its effectiveness. When tested on the QM9 dataset, comprising small organic molecules, the proposed model demonstrated superior accuracy compared to both the original EGNN models and the baseline models without PFP-derived descriptors for 11 out of the 12 molecular properties. Furthermore, when evaluated on the tmQM dataset, which encompasses transition metal complexes, notable enhancements in performance were observed across all five target properties, indicating the significance of the local atomic environment surrounding transition metals. In essence, the proposed methodology is adaptable to any 3D GNN architecture, and further enhancements in prediction accuracy are anticipated when integrated with continually evolving GNN architectures.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑