Minimally sufficient experimental design using identifiability analysis

Area of Science:

• Mathematical Biology
• Pharmacokinetics and Pharmacodynamics
• Experimental Design

Background:

• Mathematical models require data for calibration and real-time prediction.
• Optimizing data collection is crucial for enhancing model predictive power and utility.
• Parameter identifiability is key for reliable model predictions.

Purpose of the Study:

• To develop a framework for optimal experimental design to ensure parameter identifiability.
• To identify the minimal data collection strategy that maximizes model informativeness.
• To minimize experimental time and costs in mathematical model calibration.

Main Methods:

• Defining model-informative data based on unique parameterization and practical identifiability.
• Proposing a framework to determine optimal data collection timing and quantity.
• Applying the method to a pharmacokinetic/pharmacodynamic model of drug distribution in the tumor microenvironment (TME).

Main Results:

• Identified a methodology for optimal experimental design.
• Demonstrated the framework's application to a TME drug distribution model.
• Determined a minimal set of time points for robust parameter identifiability.

Conclusions:

• The proposed framework ensures practical identifiability for mathematical models.
• It enables the identification of minimally sufficient experimental designs.
• This approach minimizes experimental costs and time while maximizing data utility.