What question did this study set out to answer?

This research aims to develop a machine learning framework to predict cardiac toxicity associated with ion channel blockers.

March 26, 2026

Graph-Based Classification with GNN-Explainer for Predicting Cardiac Toxicity Associated with Multi-Ion Channel Blockers

Key Result

CardiotoxPred, a deep learning framework integrating data across Kv11.1, Cav1.2, and Nav1.5 ion channels, achieved an average prediction accuracy of 86.7% for assessing cardiotoxicity risk.

Key Points

This research aims to develop a machine learning framework to predict cardiac toxicity associated with ion channel blockers.
Developed a deep learning framework integrating inhibition data for Kv11.1, Cav1.2, and Nav1.5 ion channels.
Curated a comprehensive dataset comprising 34,124 molecules for Kv11.1 and smaller sets for Cav1.2 and Nav1.5.
Trained GNN models for individual channel predictions and used GNNExplainer for interpretability.
Achieved an average prediction accuracy of 86.7% on the test data set.
Provided visualizations of atomic and bond-level contributions for better understanding of toxicity.
Models are accessible via GitHub, enhancing practical use for drug discovery.

Structured PICO

Population

Molecules from PubChem, CUPID, and CToxPred2 repositories (34,124 for Kv11.1, 1564 for Cav1.2, and 3217 for Nav1.5)

Intervention

CardiotoxPred method (deep learning framework using Graph Neural Network models)

Outcome

Prediction accuracy of cardiac ion channel inhibition

A novel deep learning framework integrating data across multiple cardiac ion channels provides high accuracy in predicting drug-induced cardiotoxicity, offering a valuable tool for early-stage drug discovery.

Main Result

Absolute Event Rate: 0% vs 0%

Abstract

Cardiotoxicity remains a critical concern in drug development, often leading to late-stage attrition of promising compounds. While traditional assessments focus on Kv11.1 channel inhibition, the Comprehensive in Vitro Proarrhythmic Assay (CiPA) initiative emphasizes the importance of evaluating additional cardiac ion channels, notably Cav1.2 and Nav1.5. In this study, we address the limitations of existing machine learning (ML) models, which typically rely on Kv11.1-specific data, by developing a deep learning (DL) framework that integrates inhibition data across all three key ion channels. A large and diverse data set (Cardio-Tox) was curated by combining experimental data from the PubChem, CUPID, and CToxPred2 repositories, totaling 34,124 molecules for Kv11.1, 1564 for Cav1.2, and 3217 for Nav1.5. Using this data set, trained GNN models are capable of individual channel prediction. The developed CardiotoxPred method, which includes the Kv, Cav, and Nav models, achieved an average prediction accuracy of 86.7% on a test data set. In addition to robust predictive performance, GNNExplainer offers interpretable visualizations by highlighting atom- and bond-level contributions via colors. These insights support cardiac molecular severity estimation, optimization, and safety profiling. All the models are freely accessible via GitHub in a user-friendly Docker container, providing a practical tool for early-stage cardiotoxicity risk assessment in drug discovery pipelines.

Bookmark