A Bayesian framework for systems model refinement and selection of calcium signaling.

Area of Science:

• Cellular Biology
• Computational Biology
• Biophysics

Background:

• Calcium (Ca²⁺) is a vital intracellular messenger regulating cellular functions.
• Ca²⁺ signaling dysregulation is linked to diseases like cancer and heart failure.
• Existing computational models often fail to account for cell population heterogeneity.

Purpose of the Study:

• To develop an advanced statistical framework to model Ca²⁺ signaling dynamics.
• To explicitly address cell-to-cell variability and population-wide differences in Ca²⁺ signaling.
• To improve the accuracy of computational models for Ca²⁺ dynamics.

Main Methods:

• Developed a Bayesian inference framework with a hierarchical mixture architecture.
• Applied the framework to myoblasts and HEK293 cells expressing cardiac proteins.
• Utilized fluorescence microscopy to monitor Ca²⁺ dynamics and analyze cell populations.

Main Results:

• Successfully distinguished multiple clusters of cells exhibiting distinct kinetic behaviors.
• Identified probable models and parameters that accurately reproduce experimental Ca²⁺ dynamics.
• Demonstrated the framework's ability to capture and model cellular heterogeneity in Ca²⁺ signaling.

Conclusions:

• The Bayesian framework significantly enhances the accuracy of computational Ca²⁺ signaling models.
• Explicitly accounting for cellular differences improves understanding of complex Ca²⁺ regulatory networks.
• This approach offers deeper insights into biological processes and their variability across cell populations.