Bottom-up GGM algorithm for constructing multilayered hierarchical gene regulatory networks that govern biological...

Area of Science:

• Genomics
• Systems Biology
• Bioinformatics

Background:

• Multilayered hierarchical gene regulatory networks (ML-hGRNs) are crucial for understanding gene regulation in biological pathways.
• Current computational methods lack the ability to directly construct ML-hGRNs for biological pathways.

Purpose of the Study:

• To develop a novel computational algorithm for constructing ML-hGRNs.
• To enable the analysis of ML-hGRNs using small- to medium-sized gene expression datasets.

Main Methods:

• Developed a bottom-up graphic Gaussian model (GGM) algorithm.
• Constructed ML-hGRNs layer-by-layer by evaluating gene triplets and regulatory interference.
• Utilized an optimization function to select regulatory genes for subsequent layers.

Main Results:

• Successfully constructed ML-hGRNs governing biological pathways.
• The algorithm demonstrated high efficiency in handling small- to medium-sized microarray or RNA-seq data.
• Identified hierarchical regulators within biological pathways.

Conclusions:

• The developed GGM algorithm is effective for constructing ML-hGRNs.
• This tool empowers biologists to decipher hierarchical gene regulation from limited gene expression data.
• Facilitates a deeper understanding of biological pathway genetics.