Parameter estimation for gene regulatory networks: a two-stage MCMC Bayesian approach.

Area of Science:

• Systems Biology
• Computational Biology
• Biochemistry

Background:

• Genetic regulatory networks (GRNs) are crucial for understanding biological functions.
• Accurately determining the parametric forms of GRN models is a significant challenge.

Purpose of the Study:

• To present a novel computational approach for solving the inverse problem in GRN modeling.
• To provide a joint probability distribution of model parameters for more robust predictions.

Main Methods:

• A C++ implemented computational approach involving an optimization stage followed by Bayesian filtering.
• Application to time series data from gene circuit models using state space representation.

Main Results:

• The approach successfully determines parametric forms for GRN models.
• It efficiently prunes unsound terms from generalized models.
• A joint probability distribution of model parameters is generated, not just single estimates.

Conclusions:

• The developed method offers a flexible, general, and robust solution for GRN parameter estimation.
• It enhances the understanding of system biology, including behavior, mechanisms, and thermodynamics.
• This approach provides valuable insights into the dynamics of biological systems.