Many pharmacological compounds fail before reaching the market due to adverse cardiovascular effects, particularly ventricular arrhythmias linked to unwanted inhibition of hERG potassium channels, which are essential for ventricular repolarization.
Current approaches often rely on static IC50 values and steady-state dose-response curves, which do not capture the dynamic and state-dependent nature of drug–channel interactions. This can lead to false positives and negatives, increasing costs and reducing efficiency in the R&D pipeline.
To address this need, researchers from the Centro de Investigación e Innovación en Bioingeniería (Ci2B) at the Polytechnic University of Valencia have developed a computational tool to automatically and dynamically model drug–hERG/IKr channel interactions.
The technology is a computational tool that automatically generates dynamic models of drug interactions with hERG/IKr channels from electrophysiological data obtained through standard voltage-clamp protocols.
By combining machine learning and optimization methods, the system identifies drug binding and unbinding kinetics, channel-state dependence and the temporal evolution of channel blockade. The resulting model reproduces both transient and steady-state drug–channel behaviour, offering a more realistic mechanistic representation than classical IC50-based approaches.
The workflow includes the application of voltage-clamp protocols, analysis of dose-response and current inhibition over time, and automatic model generation for integration into preclinical in silico cardiotoxicity assessment.
The technology has been validated with a panel of ten reference hERG/IKr blockers used in the CiPA framework, demonstrating its ability to reproduce key dynamic features of drug–channel interactions.
The results have been supported by Nanion Technologies and published in Computer Methods and Programs in Biomedicine, reinforcing the scientific and technical robustness of the tool.
Benefits:
- Dynamic modeling captures drug–hERG/IKr interactions more accurately.
- Kinetic profiling identifies binding, unbinding and state dependence.
- Improved prediction supports better proarrhythmic risk assessment.
- Automated workflow reduces expert input and modelling time.
- CiPA alignment supports regulatory-oriented in silico evaluation.
- Preclinical integration fits standard voltage-clamp data workflows.
The represented institution seeks a collaboration that will lead to the commercial exploitation of the invention presented.
Institution: Universitat Politècnica de Valencia (UPV) & Centro de Investigación e Innovación en Biotecnología (Ci2B)
TRL: 5-6
Protection Status: Know – how
Financing: The European Union’s Horizon 2020 research and innovation program
Contact: Begoña Iborra | biborra@viromii.com