Adaptive modelling of gene regulatory network using Bayesian information criterion-guided sparse regression approach

Area of Science:

• Systems Biology
• Bioinformatics
• Computational Biology

Background:

• Inferring gene regulatory networks (GRNs) from gene expression data is crucial for understanding cellular mechanisms.
• Existing methods often struggle with the complexity and scale of biological data, leading to challenges in accurate network reconstruction.

Purpose of the Study:

• To develop an advanced computational method for reconstructing gene regulatory networks (GRNs).
• To improve the accuracy and interpretability of GRN inference from microarray expression data.

Main Methods:

• A Bayesian information criterion (BIC)-guided sparse regression approach was developed.
• The method utilizes l1-norm regularization optimized by a modified BIC to ensure sparsity and incorporate prior knowledge.
• The approach is designed to interpret combinatorial gene expression regulation.

Main Results:

• The proposed method successfully infers GRNs with high accuracy on both simulated and real-world microarray datasets.
• It effectively avoids the overestimation of gene regulators common in other methods.
• The approach provides clear interpretations of combinatorial regulatory relationships.

Conclusions:

• The BIC-guided sparse regression offers a robust and interpretable method for GRN reconstruction.
• This approach enhances the understanding of gene regulation in systems biology.
• The findings demonstrate significant advancements in analyzing gene expression data for network inference.