Calibration methods to fit parameters within complex biological models

Area of Science:

• Computational Biology
• Systems Biology
• Mathematical Modeling

Background:

• Biological models are increasingly complex, integrating multi-scale methods (ODEs, PDEs, agent-based) across biological resolutions.
• Fitting these models to experimental data is challenging due to numerous parameters and large degrees of freedom.
• Current methods struggle with fitting complex models to diverse, time- and space-resolved datasets.

Purpose of the Study:

• To review, test, and advance model calibration practices for complex biological systems.
• To compare the performance of a novel Calibration Protocol (CaliPro) against approximate Bayesian computing (ABC).
• To provide guidance on selecting appropriate calibration methods based on model and dataset characteristics.

Main Methods:

• Comparative analysis of model calibration methodologies.
• Implementation and testing of the Calibration Protocol (CaliPro) and approximate Bayesian computing (ABC).
• Evaluation of calibration performance across various model types and dataset complexities.

Main Results:

• The Calibration Protocol (CaliPro) demonstrates robust performance in calibrating complex biological models.
• Comparative analysis highlights the strengths, weaknesses, and synergies between CaliPro and ABC.
• Next-generation updates to CaliPro are presented, enhancing its capabilities.

Conclusions:

• Standardized calibration methods are crucial for advancing computational biology.
• CaliPro offers a powerful and adaptable approach for calibrating complex mechanistic biological models.
• A decision tree is proposed to guide the selection of calibration strategies for different modeling scenarios.