Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction

Abstract Fast and accurate spectral prediction plays a crucial role in molecular design within fields such as pharmaceutical and materials science. Nevertheless, predicting molecular spectra typically requires quantum chemistry calculations, posing significant challenges for fast predictions and hig...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuzhi Xu, Daqian Bian, Cheng-Wei Ju, Fanyu Zhao, Pujun Xie, Yuanqing Wang, Wei Hu, Zhenrong Sun, John Z. H. Zhang, Tong Zhu
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:npj Computational Materials
Online Access:https://doi.org/10.1038/s41524-025-01698-z
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849402039885365248
author Yuzhi Xu
Daqian Bian
Cheng-Wei Ju
Fanyu Zhao
Pujun Xie
Yuanqing Wang
Wei Hu
Zhenrong Sun
John Z. H. Zhang
Tong Zhu
author_facet Yuzhi Xu
Daqian Bian
Cheng-Wei Ju
Fanyu Zhao
Pujun Xie
Yuanqing Wang
Wei Hu
Zhenrong Sun
John Z. H. Zhang
Tong Zhu
author_sort Yuzhi Xu
collection DOAJ
description Abstract Fast and accurate spectral prediction plays a crucial role in molecular design within fields such as pharmaceutical and materials science. Nevertheless, predicting molecular spectra typically requires quantum chemistry calculations, posing significant challenges for fast predictions and high-throughput screening. In this paper, we propose an equivariant, fast, and robust model, named EnviroDetaNet, which integrates molecular environment information. EnviroDetaNet employs an E(3)-equivariant message-passing neural network combining intrinsic atomic properties, spatial features, and environmental information, allowing it to comprehensively capture both local and global molecular information. Compared to state-of-the-art machine learning models, EnviroDetaNet excels in various predictive tasks and maintains high accuracy even with a 50% reduction in training data, demonstrating strong generalization capabilities. Ablation studies confirm that molecular environment information is crucial for improving model stability and accuracy. EnviroDetaNet also shows outstanding performance in spectral predictions for complex molecular systems, making it a powerful tool for accelerating molecular discovery.
format Article
id doaj-art-07c9f95d205044efac95ebf1e4bac7d9
institution Kabale University
issn 2057-3960
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series npj Computational Materials
spelling doaj-art-07c9f95d205044efac95ebf1e4bac7d92025-08-20T03:37:38ZengNature Portfolionpj Computational Materials2057-39602025-07-0111111010.1038/s41524-025-01698-zPretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra predictionYuzhi Xu0Daqian Bian1Cheng-Wei Ju2Fanyu Zhao3Pujun Xie4Yuanqing Wang5Wei Hu6Zhenrong Sun7John Z. H. Zhang8Tong Zhu9Shanghai Engineering Research Center of Molecular Therapeutics and New Drug Development, School of Chemistry and Molecular Engineering, East China Normal UniversityDepartment of Chemistry and Pritzker School of Molecular Engineering, The University of ChicagoDepartment of Chemistry and Pritzker School of Molecular Engineering, The University of ChicagoShanghai Engineering Research Center of Molecular Therapeutics and New Drug Development, School of Chemistry and Molecular Engineering, East China Normal UniversityDepartment of Chemistry, New York UniversityDepartment of Chemistry, New York UniversitySchool of Chemistry and Chemical Engineering, Qilu University of Technology (Shandong Academy of Science)State Key Laboratory of Precision Spectroscopy, and School of Physics and Electron Science, East China Normal UniversityShanghai Engineering Research Center of Molecular Therapeutics and New Drug Development, School of Chemistry and Molecular Engineering, East China Normal UniversityShanghai Engineering Research Center of Molecular Therapeutics and New Drug Development, School of Chemistry and Molecular Engineering, East China Normal UniversityAbstract Fast and accurate spectral prediction plays a crucial role in molecular design within fields such as pharmaceutical and materials science. Nevertheless, predicting molecular spectra typically requires quantum chemistry calculations, posing significant challenges for fast predictions and high-throughput screening. In this paper, we propose an equivariant, fast, and robust model, named EnviroDetaNet, which integrates molecular environment information. EnviroDetaNet employs an E(3)-equivariant message-passing neural network combining intrinsic atomic properties, spatial features, and environmental information, allowing it to comprehensively capture both local and global molecular information. Compared to state-of-the-art machine learning models, EnviroDetaNet excels in various predictive tasks and maintains high accuracy even with a 50% reduction in training data, demonstrating strong generalization capabilities. Ablation studies confirm that molecular environment information is crucial for improving model stability and accuracy. EnviroDetaNet also shows outstanding performance in spectral predictions for complex molecular systems, making it a powerful tool for accelerating molecular discovery.https://doi.org/10.1038/s41524-025-01698-z
spellingShingle Yuzhi Xu
Daqian Bian
Cheng-Wei Ju
Fanyu Zhao
Pujun Xie
Yuanqing Wang
Wei Hu
Zhenrong Sun
John Z. H. Zhang
Tong Zhu
Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction
npj Computational Materials
title Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction
title_full Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction
title_fullStr Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction
title_full_unstemmed Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction
title_short Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction
title_sort pretrained e 3 equivariant message passing neural networks with multi level representations for organic molecule spectra prediction
url https://doi.org/10.1038/s41524-025-01698-z
work_keys_str_mv AT yuzhixu pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT daqianbian pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT chengweiju pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT fanyuzhao pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT pujunxie pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT yuanqingwang pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT weihu pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT zhenrongsun pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT johnzhzhang pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction
AT tongzhu pretrainede3equivariantmessagepassingneuralnetworkswithmultilevelrepresentationsfororganicmoleculespectraprediction