Identification of gene regulation models from single-cell data.

Area of Science:

• Systems Biology
• Computational Biology
• Molecular Biology

Background:

• Quantitative biological analyses employ diverse models (spatial/non-spatial, deterministic/stochastic) and data-fitting approaches.
• Discrepancies in results can arise even with identical data and models due to analytical choices.

Purpose of the Study:

• To illustrate how different modeling and analysis strategies affect quantitative biological insights.
• To explore parameter uncertainty in gene regulation models across various computational approaches.

Main Methods:

• Utilized a simplified gene regulation model with simulated single-cell data.
• Applied ordinary differential equation (ODE) analyses, chemical master equation (CME), finite state projection (FSP), and stochastic simulations.
• Employed Matlab and Python for analyses, considering time-dependent input signals and multiple model hypotheses.

Main Results:

• Demonstrated that deterministic and stochastic approaches yield different mechanistic and parameter insights from the same data.
• Showcased how parameter uncertainty varies significantly based on the chosen analysis method and experimental design.

Conclusions:

• Emphasize the critical impact of model selection and analysis methodology on biological process quantification.
• Advocate for careful consideration of computational approaches to ensure robust interpretation of experimental data in systems biology.