Related Experiment Video
Updated: Jun 4, 2025
Advanced 3D Liver Models for In vitro Genotoxicity Testing Following Long-Term Nanomaterial Exposure
Published on: June 5, 2020
Machine learning enhances genotoxicity assessment using MultiFlow® DNA damage assay
Panuwat Trairatphisan1, Lena Dorsheimer1, Peter Monecke1
1Research and Development, Preclinical Safety, Sanofi, Industriepark Hoechst, Frankfurt am Main, Germany.
Abstract:
Genotoxicity is a critical determinant for assessing the safety of pharmaceutical drugs, their metabolites, and impurities. Among genotoxicity tests, mechanistic assays such as the MultiFlow® DNA damage assay (MFA) allows the investigations on mode of action (MoA) of DNA damage through four mechanistic markers recorded at two time points. Previous studies have shown that machine learning (ML) can enhance precision on classifying the MoA of genotoxicants. Nevertheless, these approaches need to be tailored to specific chemical spaces and lab conditions for accurate risk assessment. In this study, we applied various state-of-the-art ML algorithms available in an open-source R package (caret) to build MFA-ML models using data from Bryce et al. (2016). The best model achieved 95% accuracy on the training dataset and correctly predicted genotoxicity in 16 out of 17 cases in the test dataset. Incorporating molecular descriptors properties from established in silico models demonstrated further improved performance of the approach to cover challenging examples of pharmaceuticals exhibiting a pharmacological mode of action that could interfere with the biomarker response. Further model validation on an external test set with 49 non-overlapped compounds showed a high model accuracy at 92%. Additionally, a tailored graphical user interface was developed using a freely available R package (shiny) to support visual analysis of MFA data including MoA predictions, facilitating broad usage by laboratory scientists. Lastly, a perspective on the integration of MoA predictions as additional evidence into a genotoxicity assessment workflow is proposed.
Insights
Machine learning models accurately predict the mode of action for genotoxicity using MultiFlow® DNA damage assay (MFA) data. This approach enhances pharmaceutical safety assessments by improving the precision of genotoxicity testing.
Area of Science:
- Pharmacology and Toxicology
- Computational Chemistry
- Biotechnology
Background:
- Genotoxicity testing is crucial for pharmaceutical safety assessment.
- Mechanistic assays like the MultiFlow® DNA damage assay (MFA) provide insights into DNA damage pathways.
- Machine learning (ML) has shown potential in enhancing the classification of genotoxicant modes of action (MoA).
Purpose of the Study:
- To develop and validate ML models for predicting the MoA of genotoxicity using MFA data.
- To improve the accuracy and reliability of genotoxicity risk assessment for pharmaceuticals.
- To create a user-friendly tool for analyzing MFA data and MoA predictions.
Main Methods:
- Application of state-of-the-art ML algorithms from the R package 'caret' to MFA data.
- Integration of molecular descriptors from in silico models to enhance model performance.
- Development of a graphical user interface using the R package 'shiny' for data visualization and analysis.
- Validation of models on training, internal test, and external test datasets.
Main Results:
- The best ML model achieved 95% accuracy on the training dataset and correctly predicted genotoxicity in 16 out of 17 cases in the test dataset.
- Incorporating molecular descriptors improved performance, particularly for complex pharmaceutical cases.
- External validation on 49 compounds demonstrated high model accuracy at 92%.
- A user-friendly graphical interface was developed to facilitate broad laboratory use.
Conclusions:
- ML models, when tailored, can significantly enhance the precision of MoA determination in genotoxicity testing.
- The developed approach offers a robust method for genotoxicity assessment, aiding in pharmaceutical safety evaluations.
- Integration of MoA predictions can serve as valuable evidence in regulatory genotoxicity assessment workflows.
