Related Experiment Video
Updated: Dec 11, 2025
A High-throughput Assay for the Prediction of Chemical Toxicity by Automated Phenotypic Profiling of Caenorhabditis elegans
Published on: March 14, 2019
Predicting nanotoxicity by an integrated machine learning and metabolomics approach.
Ting Peng1, Changhong Wei1, Fubo Yu1
1Key Laboratory of Pollution Processes and Environmental Criteria (Ministry of Education)/Tianjin Key Laboratory of Environmental Remediation and Pollution Control, College of Environmental Science and Engineering, Nankai University, Tianjin, 300350, China.
Predicting nanoparticle health risks is crucial. Machine learning models accurately forecast metabolic pathway disturbances from engineered nanoparticles (ENPs), identifying type and size as key factors for environmental health assessments.
Area of Science:
- Environmental Science
- Toxicology
- Computational Biology
Background:
- Assessing environmental health risks of engineered nanoparticles (ENPs) is vital.
- Predicting biological responses, specifically metabolic pathway disturbances, is challenging due to complex biological systems and diverse ENP properties.
Purpose of the Study:
- To develop accurate predictive models for metabolic pathway disturbances induced by ENPs.
- To identify key ENP properties influencing biological responses.
- To provide a rapid assessment method for ENP environmental health risks.
Main Methods:
- Integration of multiple machine learning models with metabolomics data.
- Screening of nine ENP properties to identify critical features.
- Application of similarity network analysis and decision tree models.
- Validation through animal experiments.
Main Results:
- Accurate prediction of metabolic pathway disturbances for 33 ENPs.
- Identification of ENP type and size as the most influential properties.
- Achieved 75%-100% model accuracy, including predictions for ENPs not in the training database.
- Successful prediction of metabolic pathway-related histopathology.
Conclusions:
- Machine learning and metabolomics offer a robust approach for predicting ENP-induced biological effects.
- ENP type and size are primary determinants of metabolic pathway disruption.
- This methodology enables efficient environmental health risk assessment for both known and novel ENPs.
