A Bayesian approach to modelling heterogeneous calcium responses in cell populations

Area of Science:

• Cellular Biology
• Computational Biology
• Biophysics

Background:

• Calcium responses are crucial for cellular signaling, translating external stimuli into internal actions.
• Existing models lack a comprehensive theory to predict the precise timing of calcium spike sequences.

Purpose of the Study:

• To develop and validate a quantitative modeling framework for predicting calcium spike sequences.
• To assess the efficacy of Gaussian processes in modeling calcium spike rates.

Main Methods:

• Utilized a Bayesian approach with Gaussian processes to model calcium spike rates.
• Compared Gaussian process performance against peri-stimulus time histograms and kernel smoothing.
• Analyzed calcium spike data from dynamically-stimulated HEK293T cells.

Main Results:

• Gaussian processes demonstrated high fidelity in modeling calcium spike rates.
• The proposed framework outperformed standard tools like peri-stimulus time histograms and kernel smoothing.
• Gaussian processes successfully described calcium spike rates despite single-cell variability and limited spike data.

Conclusions:

• Gaussian processes provide a robust method for quantitatively describing calcium spike timing.
• This modeling approach is effective even with heterogeneous cellular responses and dynamic stimuli.
• The findings advance the statistical understanding of calcium oscillations in complex biological environments.