Related Experiment Video
Updated: Aug 28, 2025
High-Throughput Cardiotoxicity Screening Using Mature Human Induced Pluripotent Stem Cell-Derived Cardiomyocyte Monolayers
Published on: March 24, 2023
Ligand-based prediction of hERG-mediated cardiotoxicity based on the integration of different machine learning
Pietro Delre1,2, Giovanna J Lavado3, Giuseppe Lamanna1,2
1CNR-Institute of Crystallography, Bari, Italy.
Insights
Predicting drug-induced cardiotoxicity is crucial for drug safety. This study developed highly predictive computational models for human ether-a-go-go-related (hERG) channel blockers using machine learning and curated chemical data.
Area of Science:
- Computational chemistry
- Drug discovery
- Toxicology
Background:
- Drug-induced cardiotoxicity is a significant safety concern, often linked to off-target interactions with the cardiac hERG potassium channel.
- Early identification of hERG channel blocking potential is essential in preclinical drug development to mitigate cardiotoxicity risks.
Purpose of the Study:
- To develop and validate robust, ligand-based computational models for predicting hERG channel-mediated cardiotoxicity.
- To establish a reliable computational workflow for assessing drug candidates' hERG blocking potential.
Main Methods:
- Trained and validated 30 ligand-based classifiers using 7,963 compounds from the ChEMBL database (version 25).
- Employed machine learning algorithms including random forest, k-nearest neighbors, gradient boosting, extreme gradient boosting, multilayer perceptron, and support vector machine.
- Utilized best practices for data curation, VSURF for feature selection, and SMOTE for handling imbalanced data.
Main Results:
- Achieved highly predictive models with maximal balanced accuracy (BA_MAX) of 0.91 and maximal area under the curve (AUC_MAX) of 0.95.
- Temporal validation demonstrated the predictivity and superior performance of the developed classifiers compared to existing literature models.
- A novel computational workflow for building predictive models of hERG-mediated cardiotoxicity was developed and made publicly available.
Conclusions:
- The developed computational models and workflow provide a valuable tool for accurately predicting hERG-mediated cardiotoxicity in early drug discovery.
- The study highlights the importance of rigorous data handling, feature selection, and imbalance management in building reliable predictive toxicology models.
- The freely available computational workflow can aid researchers in prioritizing safer drug candidates and advancing drug discovery programs.
Abstract:
Drug-induced cardiotoxicity is a common side effect of drugs in clinical use or under postmarket surveillance and is commonly due to off-target interactions with the cardiac human-ether-a-go-go-related (hERG) potassium channel. Therefore, prioritizing drug candidates based on their hERG blocking potential is a mandatory step in the early preclinical stage of a drug discovery program. Herein, we trained and properly validated 30 ligand-based classifiers of hERG-related cardiotoxicity based on 7,963 curated compounds extracted by the freely accessible repository ChEMBL (version 25). Different machine learning algorithms were tested, namely, random forest, K-nearest neighbors, gradient boosting, extreme gradient boosting, multilayer perceptron, and support vector machine. The application of 1) the best practices for data curation, 2) the feature selection method VSURF, and 3) the synthetic minority oversampling technique (SMOTE) to properly handle the unbalanced data, allowed for the development of highly predictive models (BAMAX = 0.91, AUCMAX = 0.95). Remarkably, the undertaken temporal validation approach not only supported the predictivity of the herein presented classifiers but also suggested their ability to outperform those models commonly used in the literature. From a more methodological point of view, the study put forward a new computational workflow, freely available in the GitHub repository (https://github.com/PDelre93/hERG-QSAR), as valuable for building highly predictive models of hERG-mediated cardiotoxicity.
More Related Videos
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
10:30Technical Applications of Microelectrode Array and Patch Clamp Recordings on Human Induced Pluripotent Stem Cell-Derived Cardiomyocytes
Published on: August 4, 2022
