Boolean dynamics of genetic regulatory networks inferred from microarray time series data

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Current methods for inferring genetic regulatory networks typically aim to identify a single network by optimizing parameters to fit experimental data.
• Theoretical work suggests biological networks possess robustness and adaptability, potentially explainable by dynamical basins of attraction.

Purpose of the Study:

• To investigate the inference of multiple genetic regulatory networks that produce similar system dynamics.
• To explore the implications of network robustness and adaptability in genetic regulation.

Main Methods:

• Developed a novel method for inferring genetic regulatory networks from time series microarray data.
• Employed k-means clustering and support vector regression for gene expression data discretization.
• Enumerated Boolean activation-inhibition networks and analyzed their dynamics.

Main Results:

• Successfully applied the method to two immunology datasets (T cell and macrophage responses).
• Discovered that numerous Boolean networks could match the experimental data.
• Observed that the majority of these inferred networks demonstrated similar dynamical behaviors.

Conclusions:

• The inference of multiple, dynamically similar genetic regulatory networks is feasible.
• This approach provides insights into the inherent robustness and adaptability of biological systems.
• The findings support the concept of dynamical basins of attraction in understanding gene regulatory network behavior.