Inference of Gene Regulatory Networks Using Bayesian Nonparametric Regression and Topology Information.

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Gene regulatory networks (GRNs) are crucial for understanding cellular functions and biological processes.
• Inferring GRNs from gene expression data is challenging due to data limitations and complexity.
• Existing inference algorithms often neglect network topology, limiting their accuracy.

Purpose of the Study:

• To develop an advanced Bayesian method for inferring gene regulatory networks.
• To enhance GRN inference by integrating gene expression data with prior network topology information.
• To improve gene selection and parameter estimation in nonparametric models.

Main Methods:

• A Bayesian group lasso with spike and slab priors was employed for gene selection and estimation.
• Nonparametric models utilizing B-spline basis functions were used to capture nonlinear gene relationships.
• Network topology information was incorporated as a prior within the Bayesian framework.
• The method was validated on benchmark datasets (DREAM3, DREAM4) and real biological data.

Main Results:

• The proposed Bayesian method demonstrated superior performance compared to existing GRN inference algorithms.
• Incorporating topology information as a prior significantly improved the accuracy of GRN inference.
• The use of B-spline basis functions effectively modeled nonlinear gene interactions.
• The method successfully performed gene selection and parameter estimation.

Conclusions:

• The developed Bayesian approach offers a robust and accurate method for inferring gene regulatory networks.
• Integrating network topology information is a valuable strategy for enhancing GRN inference accuracy.
• This method provides a powerful tool for systems biology research and understanding complex biological systems.