Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysis

Summary: Optimization of metal-ligand asymmetric catalysts is usually done by empirical trials, where the ligand is arbitrarily modified, and the new catalyst is re-evaluated in the lab. This procedure is not efficient and alternative strategies are highly desirable. We propose the Homogeneous Catal...

Full description

Saved in:
Bibliographic Details
Main Authors: Eduardo Aguilar-Bejarano, Ender Özcan, Raja K. Rit, Hongyi Li, Hon Wai Lam, Jonathan C. Moore, Simon Woodward, Grazziela Figueredo
Format: Article
Language:English
Published: Elsevier 2025-03-01
Series:iScience
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2589004225001415
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850190065346543616
author Eduardo Aguilar-Bejarano
Ender Özcan
Raja K. Rit
Hongyi Li
Hon Wai Lam
Jonathan C. Moore
Simon Woodward
Grazziela Figueredo
author_facet Eduardo Aguilar-Bejarano
Ender Özcan
Raja K. Rit
Hongyi Li
Hon Wai Lam
Jonathan C. Moore
Simon Woodward
Grazziela Figueredo
author_sort Eduardo Aguilar-Bejarano
collection DOAJ
description Summary: Optimization of metal-ligand asymmetric catalysts is usually done by empirical trials, where the ligand is arbitrarily modified, and the new catalyst is re-evaluated in the lab. This procedure is not efficient and alternative strategies are highly desirable. We propose the Homogeneous Catalyst Graph Neural Network (HCat-GNet), a machine learning model capable of aiding ligand optimization. This method trains models to predict the enantioselectivity of asymmetric reactions using only the SMILES representations of the participant molecules. HCat-GNet allows high interpretability indicating from which atoms the model gathers the most predictive information, thus showing which atoms within the ligand most affect the increase or decrease in the reaction’s selectivity. The validation of the model’s selectivity predictions is made using a new class of ligand for rhodium-catalyzed asymmetric 1,4-addition, demonstrating the ability of HCat-GNet to extrapolate into unknown chiral ligand space. Validation with other benchmark asymmetric reaction datasets demonstrates its generality when modeling different reactions.
format Article
id doaj-art-78b63a978cb944518697b7d151a8d84e
institution OA Journals
issn 2589-0042
language English
publishDate 2025-03-01
publisher Elsevier
record_format Article
series iScience
spelling doaj-art-78b63a978cb944518697b7d151a8d84e2025-08-20T02:15:24ZengElsevieriScience2589-00422025-03-0128311188110.1016/j.isci.2025.111881Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysisEduardo Aguilar-Bejarano0Ender Özcan1Raja K. Rit2Hongyi Li3Hon Wai Lam4Jonathan C. Moore5Simon Woodward6Grazziela Figueredo7GSK Carbon Neutral Laboratories for Sustainable Chemistry, University of Nottingham, Jubilee Campus, Triumph Road, Nottingham NG7 2TU, UK; School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UK; School of Computer Science, University of Nottingham, Jubilee Campus, Wollaton Road, Nottingham, Nottingham NG8 1BB, UKSchool of Computer Science, University of Nottingham, Jubilee Campus, Wollaton Road, Nottingham, Nottingham NG8 1BB, UKGSK Carbon Neutral Laboratories for Sustainable Chemistry, University of Nottingham, Jubilee Campus, Triumph Road, Nottingham NG7 2TU, UK; School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UKGSK Carbon Neutral Laboratories for Sustainable Chemistry, University of Nottingham, Jubilee Campus, Triumph Road, Nottingham NG7 2TU, UK; School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UKGSK Carbon Neutral Laboratories for Sustainable Chemistry, University of Nottingham, Jubilee Campus, Triumph Road, Nottingham NG7 2TU, UK; School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UKGSK Carbon Neutral Laboratories for Sustainable Chemistry, University of Nottingham, Jubilee Campus, Triumph Road, Nottingham NG7 2TU, UK; School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UKGSK Carbon Neutral Laboratories for Sustainable Chemistry, University of Nottingham, Jubilee Campus, Triumph Road, Nottingham NG7 2TU, UK; School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UK; Corresponding authorSchool of Medicine, University of Nottingham, Medical School, Nottingham NG7 2UH, UK; Corresponding authorSummary: Optimization of metal-ligand asymmetric catalysts is usually done by empirical trials, where the ligand is arbitrarily modified, and the new catalyst is re-evaluated in the lab. This procedure is not efficient and alternative strategies are highly desirable. We propose the Homogeneous Catalyst Graph Neural Network (HCat-GNet), a machine learning model capable of aiding ligand optimization. This method trains models to predict the enantioselectivity of asymmetric reactions using only the SMILES representations of the participant molecules. HCat-GNet allows high interpretability indicating from which atoms the model gathers the most predictive information, thus showing which atoms within the ligand most affect the increase or decrease in the reaction’s selectivity. The validation of the model’s selectivity predictions is made using a new class of ligand for rhodium-catalyzed asymmetric 1,4-addition, demonstrating the ability of HCat-GNet to extrapolate into unknown chiral ligand space. Validation with other benchmark asymmetric reaction datasets demonstrates its generality when modeling different reactions.http://www.sciencedirect.com/science/article/pii/S2589004225001415Artificial intelligenceCatalysisChemistry
spellingShingle Eduardo Aguilar-Bejarano
Ender Özcan
Raja K. Rit
Hongyi Li
Hon Wai Lam
Jonathan C. Moore
Simon Woodward
Grazziela Figueredo
Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysis
iScience
Artificial intelligence
Catalysis
Chemistry
title Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysis
title_full Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysis
title_fullStr Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysis
title_full_unstemmed Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysis
title_short Homogeneous catalyst graph neural network: A human-interpretable graph neural network tool for ligand optimization in asymmetric catalysis
title_sort homogeneous catalyst graph neural network a human interpretable graph neural network tool for ligand optimization in asymmetric catalysis
topic Artificial intelligence
Catalysis
Chemistry
url http://www.sciencedirect.com/science/article/pii/S2589004225001415
work_keys_str_mv AT eduardoaguilarbejarano homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis
AT enderozcan homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis
AT rajakrit homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis
AT hongyili homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis
AT honwailam homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis
AT jonathancmoore homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis
AT simonwoodward homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis
AT grazzielafigueredo homogeneouscatalystgraphneuralnetworkahumaninterpretablegraphneuralnetworktoolforligandoptimizationinasymmetriccatalysis