We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Predicting Dose-Range Chemical Toxicity using Novel Hybrid Deep Machine-Learning Method
Summary
Researchers developed a hybrid neural network deep learning method called HNN-Tox that predicts chemical toxicity at different doses by combining convolutional and fully connected neural networks, offering a faster alternative to traditional in vivo toxicity testing.
Humans are exposed to thousands of chemicals, including environmental chemicals. Unfortunately, little is known about their potential toxicity, as determining the toxicity remains challenging due to the substantial resources required to assess a chemical in vivo. Here, we present a novel hybrid neural network (HNN) deep learning method, called HNN-Tox, to predict chemical toxicity at different doses. To develop a hybrid HNN-Tox method, we combined two neural network frameworks, the Convolutional Neural Network (CNN) and the multilayer perceptron (MLP)-type feed-forward neural network (FFNN). Combining the CNN and FCNN in the field of environmental chemical toxicity prediction is a novel approach. We developed several binary and multiclass classification models to assess dose-range chemical toxicity that is trained based on thousands of chemicals with known toxicity. The performance of the HNN-Tox was compared with other machine-learning methods, including Random Forest (RF), Bootstrap Aggregation (Bagging), and Adaptive Boosting (AdaBoost). We also analyzed the model performance dependency on varying features, descriptors, dataset size, route of exposure, and toxic dose. The HNN-Tox model, trained on 59,373 chemicals annotated with known LD50 and routes of exposure, maintained its predictive ability with an accuracy of 84.9% and 84.1%, even after reducing the descriptor size from 318 to 51, and the area under the ROC curve (AUC) was 0.89 and 0.88, respectively. Further, we validated the HNN-Tox with several external toxic chemical datasets on a large scale. The HNN-Tox performed optimally or better than the other machine-learning methods for diverse chemicals. This study is the first to report a large-scale prediction of dose-range chemical toxicity with varying features. The HNN-Tox has broad applicability in predicting toxicity for diverse chemicals and could serve as an alternative methodology approach to animal-based toxicity assessment.
Sign in to start a discussion.
More Papers Like This
A Novel Hybrid IOT Based Artificial Intelligence Algorithm for Toxicity Prediction In The Environment And Its Effect On Human Health
Researchers proposed a hybrid IoT-based artificial intelligence framework for predicting environmental toxicity and its effects on human health, combining sensor networks with machine learning to improve real-time assessment of chemical exposure risks in the environment.
Machine learning-driven QSAR models for predicting the mixture toxicity of nanoparticles
Researchers used machine learning to predict how toxic different mixtures of metal nanoparticles are to bacteria. Their models outperformed traditional methods at predicting combined toxicity effects. While focused on engineered nanoparticles rather than microplastics, the computational approach could be adapted to predict health risks from the complex mixtures of nano-sized pollutants people encounter.
Predicting the toxicity of microplastic particles through machine learning models
Researchers developed machine learning models to predict microplastic particle toxicity from physical and chemical descriptors, addressing the classification challenge posed by the enormous diversity of particle types that cannot be characterized using conventional chemical hazard methods. The models provided accurate toxicity predictions across diverse microplastic types, offering a practical screening tool for the field.
A Screening Method to Identify Potential Endocrine Disruptors Using Chemical Toxicity Big Data and a Deep Learning Model with a Focus on Cleaning and Laundry Products
Researchers used toxicity databases and deep learning models to screen cleaning and laundry products for potential endocrine-disrupting chemicals (EDCs). The model identified several compounds in common household products that are likely EDCs, many of which are also associated with plastic packaging or plastic-based product formulations. This approach could accelerate the identification of hazardous chemicals in consumer products without animal testing.
Transfer learning enables robust prediction of cellular toxicity from environmental micro- and nanoplastics
Researchers developed a transfer learning approach to predict cellular toxicity from micro- and nanoplastics, overcoming the challenge of limited experimental data. By pre-training a model on a large nanoparticle dataset and fine-tuning it on plastic-specific data, they achieved strong predictive accuracy. The tool allows researchers to estimate the toxicity of various plastic particles based on their physical and chemical properties without extensive new experiments.