Conf-GEM: A geometric information-assisted direct conformation generation model

Molecular conformations generation (MCG) aims to efficiently obtain reasonable and stable three-dimensional (3D) atomic coordinates of the atoms in the molecule from scratch, providing a structural foundation for molecular representation learning models and advanced downstream molecular design tasks...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhijiang Yang, Youjun Xu, Li Pan, Tengxin Huang, Yunfan Wang, Junjie Ding, Liangliang Wang, Junhua Xiao
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Artificial Intelligence Chemistry
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2949747724000320
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846137552726130688
author Zhijiang Yang
Youjun Xu
Li Pan
Tengxin Huang
Yunfan Wang
Junjie Ding
Liangliang Wang
Junhua Xiao
author_facet Zhijiang Yang
Youjun Xu
Li Pan
Tengxin Huang
Yunfan Wang
Junjie Ding
Liangliang Wang
Junhua Xiao
author_sort Zhijiang Yang
collection DOAJ
description Molecular conformations generation (MCG) aims to efficiently obtain reasonable and stable three-dimensional (3D) atomic coordinates of the atoms in the molecule from scratch, providing a structural foundation for molecular representation learning models and advanced downstream molecular design tasks such as molecular property prediction, molecular generation, and molecular docking. Existing MCG methods mostly rely on indirect distance-based strategies, which which can result in geometrically unrealistic conformations, or direct coordinate-based methods, which have larger search spaces and are prone to overfitting. Therefore, this study introduces Conf-GEM, a novel geometric information-assisted direct conformation generation model based on E-GeoGNN, a geometrically augmented 3D graph neural network with multiple scales. Pre-training and divide-and-conquer strategies, are integrated into the proposed model. Conf-GEM outperforms RDKit and nine deep-learning-based MCG models on the GEOM-QM9 and GEOM-Drugs datasets, achieving conformational coverage of 96.69% and 96.07%, respectively, without force field optimization. It also excels on the X-ray diffraction crystal structure dataset with up to 97.04% conformational coverage. In conclusion, Conf-GEM provides a novel solution for stabilizing 3D conformations generation. We provide an online prediction service (https://confgem.cmdrg.com) with a user-friendly interface for researchers.
format Article
id doaj-art-0773a5801972491bacf28671f3bf0f96
institution Kabale University
issn 2949-7477
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Artificial Intelligence Chemistry
spelling doaj-art-0773a5801972491bacf28671f3bf0f962024-12-08T06:13:44ZengElsevierArtificial Intelligence Chemistry2949-74772024-12-0122100074Conf-GEM: A geometric information-assisted direct conformation generation modelZhijiang Yang0Youjun Xu1Li Pan2Tengxin Huang3Yunfan Wang4Junjie Ding5Liangliang Wang6Junhua Xiao7State Key Laboratory of NBC Protection for Civilian, ChinaInfinite Intelligence Pharma, Beijing 100020, ChinaState Key Laboratory of NBC Protection for Civilian, ChinaState Key Laboratory of NBC Protection for Civilian, ChinaState Key Laboratory of NBC Protection for Civilian, ChinaState Key Laboratory of NBC Protection for Civilian, China; Corresponding authors.State Key Laboratory of NBC Protection for Civilian, China; Corresponding authors.State Key Laboratory of NBC Protection for Civilian, China; Corresponding authors.Molecular conformations generation (MCG) aims to efficiently obtain reasonable and stable three-dimensional (3D) atomic coordinates of the atoms in the molecule from scratch, providing a structural foundation for molecular representation learning models and advanced downstream molecular design tasks such as molecular property prediction, molecular generation, and molecular docking. Existing MCG methods mostly rely on indirect distance-based strategies, which which can result in geometrically unrealistic conformations, or direct coordinate-based methods, which have larger search spaces and are prone to overfitting. Therefore, this study introduces Conf-GEM, a novel geometric information-assisted direct conformation generation model based on E-GeoGNN, a geometrically augmented 3D graph neural network with multiple scales. Pre-training and divide-and-conquer strategies, are integrated into the proposed model. Conf-GEM outperforms RDKit and nine deep-learning-based MCG models on the GEOM-QM9 and GEOM-Drugs datasets, achieving conformational coverage of 96.69% and 96.07%, respectively, without force field optimization. It also excels on the X-ray diffraction crystal structure dataset with up to 97.04% conformational coverage. In conclusion, Conf-GEM provides a novel solution for stabilizing 3D conformations generation. We provide an online prediction service (https://confgem.cmdrg.com) with a user-friendly interface for researchers.http://www.sciencedirect.com/science/article/pii/S2949747724000320Molecular conformations generationVariational autoencoderDivide-and-conquer methodPre-trainingGeometric enhanced graph neural network
spellingShingle Zhijiang Yang
Youjun Xu
Li Pan
Tengxin Huang
Yunfan Wang
Junjie Ding
Liangliang Wang
Junhua Xiao
Conf-GEM: A geometric information-assisted direct conformation generation model
Artificial Intelligence Chemistry
Molecular conformations generation
Variational autoencoder
Divide-and-conquer method
Pre-training
Geometric enhanced graph neural network
title Conf-GEM: A geometric information-assisted direct conformation generation model
title_full Conf-GEM: A geometric information-assisted direct conformation generation model
title_fullStr Conf-GEM: A geometric information-assisted direct conformation generation model
title_full_unstemmed Conf-GEM: A geometric information-assisted direct conformation generation model
title_short Conf-GEM: A geometric information-assisted direct conformation generation model
title_sort conf gem a geometric information assisted direct conformation generation model
topic Molecular conformations generation
Variational autoencoder
Divide-and-conquer method
Pre-training
Geometric enhanced graph neural network
url http://www.sciencedirect.com/science/article/pii/S2949747724000320
work_keys_str_mv AT zhijiangyang confgemageometricinformationassisteddirectconformationgenerationmodel
AT youjunxu confgemageometricinformationassisteddirectconformationgenerationmodel
AT lipan confgemageometricinformationassisteddirectconformationgenerationmodel
AT tengxinhuang confgemageometricinformationassisteddirectconformationgenerationmodel
AT yunfanwang confgemageometricinformationassisteddirectconformationgenerationmodel
AT junjieding confgemageometricinformationassisteddirectconformationgenerationmodel
AT liangliangwang confgemageometricinformationassisteddirectconformationgenerationmodel
AT junhuaxiao confgemageometricinformationassisteddirectconformationgenerationmodel