Abstract: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.

Benchmarking Toxic Molecule Classification using Graph Neural Networks and Few Shot Learning

Few-shot learning via graph embeddings with convolutional networks for low-data molecular property prediction

Chemical toxicity prediction based on semi-supervised learning and graph convolutional neural network

Equivariant Graph Neural Networks for Toxicity Prediction

Benchmarking Accuracy and Generalizability of Four Graph Neural Networks Using Large In Vitro ADME Datasets from Different Chemical Spaces

Multimodal Representation Learning via Graph Isomorphism Network for Toxicity Multitask Learning

Efficient Toxicity Prediction via Simple Features Using Shallow Neural Networks and Decision Trees

Multitask CapsNet: an Imbalanced Data Deep Learning Method for Predicting Toxicants

deepFPlearn +: enhancing toxicity prediction across the chemical universe using graph neural networks

Meta Learning With Graph Attention Networks for Low-Data Drug Discovery

A deep learning based multi-model approach for predicting drug-like chemical compound's toxicity

Few-Shot Graph Classification with Structural-Enhanced Contrastive Learning for Graph Data Copyright Protection

Structure-Enhanced Meta-Learning for Few-Shot Graph Classification

Predicting Cardiotoxicity of Molecules Using Attention-Based Graph Neural Networks

Graph Kernel Learning for Predictive Toxicity Models

Hi-MGT: A hybrid molecule graph transformer for toxicity identification

A New Fingerprint and Graph Hybrid Neural Network for Predicting Molecular Properties

TOP: Towards Better Toxicity Prediction by Deep Molecular Representation Learning

A Bayesian graph convolutional network for reliable prediction of molecular properties with uncertainty quantification

Mining Toxicity Information from Large Amounts of Toxicity Data

GraphADT: empowering interpretable predictions of acute dermal toxicity with multi-view graph pooling and structure remapping