用于含能分子性质预测的分子描述符增强图神经网络
Molecular descriptor-enhanced graph neural network for energetic molecular property prediction
高天宇 1纪玉金 1刘成 1李有勇2
作者信息
- 1. Institute of Functional Nano & Soft Materials(FUNSOM),Jiangsu Key Laboratory for Carbon-Based Functional Materials & Devices,Soochow University,Suzhou 215123,China
- 2. Institute of Functional Nano & Soft Materials(FUNSOM),Jiangsu Key Laboratory for Carbon-Based Functional Materials & Devices,Soochow University,Suzhou 215123,China;Macao Institute of Materials Science and Engineering,Macau University of Science and Technology,Macao SAR 999078,China
- 折叠
摘要
含能分子在军事和民用应用中都发挥着重要作用.传统上,确定含能分子的物理化学参数需要实验工作量且具有风险,而新兴的机器学习方法有望解决这一问题.在这项工作中,我们报道了一种分子描述符增强的图神经网络(MD增强的GNN)模型,该模型可以准确快速地预测含能分子的三个爆轰参数.该模型集成了基于序列的分子描述符和基于结构的图向量,提供了一个不需要自定义描述符的全面框架.因此,我们构建了一个包含18,991个CHNO含能分子的含能分子数据集,并将我们的模型与单一的分子指纹/描述符和GNN方法进行了比较.研究发现,我们提出的MD增强的GNN集成方法通过结合两个不同的互补特征,实现了卓越的精度,R2超过0.93,学习速度提高了20％以上,这突出了我们的模型在重塑含能分子设计格局方面的潜力,并有望在这一关键领域内大幅提高效率和有效性.
Abstract
Energetic molecules(EMs)play an important role in both military and civilian applications.Traditionally,determining the physicochemical parameters of EMs requires experimental workload and inherent risks while new-rising machine learning(ML)methods are promising to address this challenge.In this work,we report a molecular descriptor-enhanced graph neural network(MD-enhanced GNN)model to accurately and fast predict three detonation parameters of EMs.This model integrates sequence-based molecular de-scriptors and structure-based graph vectors,offering a com-prehensive framework that does not require custom descriptors.Accordingly,we construct an EMs dataset that includes 18,991 CHNO EMs and compare our model with sole molecular fingerprint/descriptor and GNN methods.It is found that our proposed MD-enhanced GNN integration method achieves superior accuracy with R2 over 0.93 and a learning speed improvement of over 20％by combining two different complementary features,which highlights the po-tential of our model in reshaping the landscape of EMs design,promising substantial improvements in both efficiency and effectiveness within this critical field.
关键词
energetic molecules/molecular descriptors/graph neural networkKey words
energetic molecules/molecular descriptors/graph neural network引用本文复制引用
基金项目
National Key Research Program of China(2022YFA1503101)
国家自然科学基金(22173067)
Science and Technology Development Fund,Macau SAR(FDCT)
Science and Technology Development Fund,Macau SAR(0024/2022/ITP)
Collaborative Innovation Center of Suzhou Nano Science and Technology()
江苏高校优势学科建设工程项目()
高等学校学科创新引智计划(111计划)()
Joint International Research Laboratory of Carbon-Based Functional Materials and Devices()
出版年
2024
NETL
NSTL
万方数据