Cargando…

Equivariant Graph Neural Networks for Toxicity Prediction

[Image: see text] Predictive modeling of toxicity is a crucial step in the drug discovery pipeline. It can help filter out molecules with a high probability of failing in the early stages of de novo drug design. Thus, several machine learning (ML) models have been developed to predict the toxicity o...

Descripción completa

Detalles Bibliográficos
Autores principales: Cremer, Julian, Medrano Sandonas, Leonardo, Tkatchenko, Alexandre, Clevert, Djork-Arné, De Fabritiis, Gianni
Formato: Online Artículo Texto
Lenguaje:English
Publicado: American Chemical Society 2023
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10583285/
https://www.ncbi.nlm.nih.gov/pubmed/37690056
http://dx.doi.org/10.1021/acs.chemrestox.3c00032
_version_ 1785122517413789696
author Cremer, Julian
Medrano Sandonas, Leonardo
Tkatchenko, Alexandre
Clevert, Djork-Arné
De Fabritiis, Gianni
author_facet Cremer, Julian
Medrano Sandonas, Leonardo
Tkatchenko, Alexandre
Clevert, Djork-Arné
De Fabritiis, Gianni
author_sort Cremer, Julian
collection PubMed
description [Image: see text] Predictive modeling of toxicity is a crucial step in the drug discovery pipeline. It can help filter out molecules with a high probability of failing in the early stages of de novo drug design. Thus, several machine learning (ML) models have been developed to predict the toxicity of molecules by combining classical ML techniques or deep neural networks with well-known molecular representations such as fingerprints or 2D graphs. But the more natural, accurate representation of molecules is expected to be defined in physical 3D space like in ab initio methods. Recent studies successfully used equivariant graph neural networks (EGNNs) for representation learning based on 3D structures to predict quantum-mechanical properties of molecules. Inspired by this, we investigated the performance of EGNNs to construct reliable ML models for toxicity prediction. We used the equivariant transformer (ET) model in TorchMD-NET for this. Eleven toxicity data sets taken from MoleculeNet, TDCommons, and ToxBenchmark have been considered to evaluate the capability of ET for toxicity prediction. Our results show that ET adequately learns 3D representations of molecules that can successfully correlate with toxicity activity, achieving good accuracies on most data sets comparable to state-of-the-art models. We also test a physicochemical property, namely, the total energy of a molecule, to inform the toxicity prediction with a physical prior. However, our work suggests that these two properties can not be related. We also provide an attention weight analysis for helping to understand the toxicity prediction in 3D space and thus increase the explainability of the ML model. In summary, our findings offer promising insights considering 3D geometry information via EGNNs and provide a straightforward way to integrate molecular conformers into ML-based pipelines for predicting and investigating toxicity prediction in physical space. We expect that in the future, especially for larger, more diverse data sets, EGNNs will be an essential tool in this domain.
format Online
Article
Text
id pubmed-10583285
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher American Chemical Society
record_format MEDLINE/PubMed
spelling pubmed-105832852023-10-19 Equivariant Graph Neural Networks for Toxicity Prediction Cremer, Julian Medrano Sandonas, Leonardo Tkatchenko, Alexandre Clevert, Djork-Arné De Fabritiis, Gianni Chem Res Toxicol [Image: see text] Predictive modeling of toxicity is a crucial step in the drug discovery pipeline. It can help filter out molecules with a high probability of failing in the early stages of de novo drug design. Thus, several machine learning (ML) models have been developed to predict the toxicity of molecules by combining classical ML techniques or deep neural networks with well-known molecular representations such as fingerprints or 2D graphs. But the more natural, accurate representation of molecules is expected to be defined in physical 3D space like in ab initio methods. Recent studies successfully used equivariant graph neural networks (EGNNs) for representation learning based on 3D structures to predict quantum-mechanical properties of molecules. Inspired by this, we investigated the performance of EGNNs to construct reliable ML models for toxicity prediction. We used the equivariant transformer (ET) model in TorchMD-NET for this. Eleven toxicity data sets taken from MoleculeNet, TDCommons, and ToxBenchmark have been considered to evaluate the capability of ET for toxicity prediction. Our results show that ET adequately learns 3D representations of molecules that can successfully correlate with toxicity activity, achieving good accuracies on most data sets comparable to state-of-the-art models. We also test a physicochemical property, namely, the total energy of a molecule, to inform the toxicity prediction with a physical prior. However, our work suggests that these two properties can not be related. We also provide an attention weight analysis for helping to understand the toxicity prediction in 3D space and thus increase the explainability of the ML model. In summary, our findings offer promising insights considering 3D geometry information via EGNNs and provide a straightforward way to integrate molecular conformers into ML-based pipelines for predicting and investigating toxicity prediction in physical space. We expect that in the future, especially for larger, more diverse data sets, EGNNs will be an essential tool in this domain. American Chemical Society 2023-09-10 /pmc/articles/PMC10583285/ /pubmed/37690056 http://dx.doi.org/10.1021/acs.chemrestox.3c00032 Text en © 2023 The Authors. Published by American Chemical Society https://creativecommons.org/licenses/by/4.0/Permits the broadest form of re-use including for commercial purposes, provided that author attribution and integrity are maintained (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Cremer, Julian
Medrano Sandonas, Leonardo
Tkatchenko, Alexandre
Clevert, Djork-Arné
De Fabritiis, Gianni
Equivariant Graph Neural Networks for Toxicity Prediction
title Equivariant Graph Neural Networks for Toxicity Prediction
title_full Equivariant Graph Neural Networks for Toxicity Prediction
title_fullStr Equivariant Graph Neural Networks for Toxicity Prediction
title_full_unstemmed Equivariant Graph Neural Networks for Toxicity Prediction
title_short Equivariant Graph Neural Networks for Toxicity Prediction
title_sort equivariant graph neural networks for toxicity prediction
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10583285/
https://www.ncbi.nlm.nih.gov/pubmed/37690056
http://dx.doi.org/10.1021/acs.chemrestox.3c00032
work_keys_str_mv AT cremerjulian equivariantgraphneuralnetworksfortoxicityprediction
AT medranosandonasleonardo equivariantgraphneuralnetworksfortoxicityprediction
AT tkatchenkoalexandre equivariantgraphneuralnetworksfortoxicityprediction
AT clevertdjorkarne equivariantgraphneuralnetworksfortoxicityprediction
AT defabritiisgianni equivariantgraphneuralnetworksfortoxicityprediction