Ion channel model reduction using manifold boundaries

Area of Science:

• Computational biology
• Biophysics
• Pharmacology

Background:

• Mathematical models of voltage-gated ion channels are crucial in research and clinical settings.
• Model complexity and structure selection remain challenging, particularly with limited data.
• The human Ether-à-go-go related gene (hERG) potassium channel is vital for cardiac electrical activity.

Purpose of the Study:

• To simplify a complex mathematical model of the cardiac hERG potassium channel.
• To investigate the efficacy of the manifold boundary approximation method (MBAM) in reducing model complexity.
• To assess the predictive capability of simplified models compared to the full model.

Main Methods:

• Application of the manifold boundary approximation method (MBAM) to an established five-state hERG model.
• Generation of a series of reduced-complexity models with fewer states and parameters.
• Validation of simplified models using experimental hERG1a data from HEK293 cells at 37°C.

Main Results:

• Reduced-complexity hERG models were successfully generated using MBAM.
• Simplified models, with up to three fewer states and eight fewer parameters, retained significant predictive capability.
• Validation confirmed the accuracy of the reduced models against experimental data.

Conclusions:

• MBAM offers an effective approach to simplify complex ion channel models.
• Model simplification improves parameter identifiability, aiding future ion channel model development.
• This method facilitates the use of more manageable and interpretable ion channel models in various applications.