Why the study?
Drug interference with cardiac ion channels leads to serious cardiovascular complications and drug discontinuations, necessitating early prediction of potential blockers during drug discovery.
Population
Curated dataset of small molecule drug candidates targeting hERG, Cav1.2, and Nav1.5 channels
Comparison
Molecular fingerprints vs descriptors vs graph-based numerical representations
Design
In silico deep learning benchmarking study
Key result
A deep learning framework, CToxPred, was developed to predict small molecule cardiotoxicity targeting hERG, Cav1.2, and Nav1.5 channels using molecular fingerprints, descriptors, and graphs.
Authors
Loading...
May aid early cardiotoxicity screening in drug discovery; leaves open prospective validation before clinical use.
CToxPred provides an open-source deep learning framework for early prediction of drug-induced cardiotoxicity targeting key cardiac ion channels.
Arab et al. (2023) studied Cardiotoxicity (cardiac ion channel inhibition). CToxPred (deep learning framework) was evaluated on Predictive capabilities of three feature representations for cardiotoxicity. A deep learning framework, CToxPred, was developed to predict small molecule cardiotoxicity targeting hERG, Cav1.2, and Nav1.5 channels using molecular fingerprints, descriptors, and graphs.