Inference of gene regulatory networks using S-system: a unified approach

Area of Science:

• Systems Biology
• Computational Biology
• Bioinformatics

Background:

• DNA microarray time-series data enables dynamic gene regulatory network (GRN) discovery.
• S-system models capture GRN dynamics but are limited by parameter complexity and data scarcity for large networks.

Purpose of the Study:

• To propose a unified approach for inferring GRNs using the S-system model.
• To enable the discovery of large-scale GRN structures and detailed modeling of small GRNs.

Main Methods:

• A simplified S-system model for fast parameter estimation and identification of major gene interactions.
• A two-step method involving genetic programming, recursive least square estimation, and multi-dimensional optimization for detailed S-system parameter estimation.

Main Results:

• The simplified S-system model demonstrated applicability on a 50-dimensional synthetic model.
• The two-step method effectively identified detailed gene interactions in small GRNs using yeast protein synthesis data.

Conclusions:

• The proposed unified approach enhances the applicability of S-system models for GRN inference.
• The methods facilitate both large-scale GRN structure discovery and detailed small-scale GRN modeling.