Parameter optimization technique using the response surface methodology

Area of Science:

• Computational Biology
• Biophysics
• Cellular Physiology

Background:

• Accurate biological cell simulation requires precise parameter values for differential equations governing cell behavior.
• Many critical parameters, such as ion permeability and ion concentration dependencies, are not directly measurable experimentally.
• Existing simulation methods struggle with parameter estimation due to the complexity and non-linear nature of biological systems.

Purpose of the Study:

• To develop and validate a robust parameter optimization technique for accurate biological cell simulations.
• To enhance the predictive power of cell models by accurately estimating unmeasurable parameters.
• To improve the simulation of cardiac myocyte action potentials, calcium transients, contraction, and ATP consumption.

Main Methods:

• Utilized Response Surface Methodology (RSM) for parameter optimization in biological cell simulations.
• Employed numerical integration of differential equations to model cell behavior.
• Implemented a recursive subdivision technique with the coefficient of multiple determination for optimizing complex parameter spaces that cannot be approximated by a single quadratic polynomial.

Main Results:

• Successfully optimized key parameters for biological cell simulations, including ion permeability and Ko dependency.
• Demonstrated the effectiveness of the proposed RSM-based technique in minimizing the difference between measured and calculated action potentials.
• Validated the method by accurately searching for pre-determined parameters in cardiac myocyte models.

Conclusions:

• The proposed parameter optimization technique using RSM provides an effective approach for accurate biological cell simulations.
• This method addresses the challenge of estimating unmeasurable parameters crucial for predicting cell electrophysiology and mechanics.
• The recursive subdivision strategy enhances RSM's applicability to complex, multi-dimensional parameter spaces in computational biology.