Large-scale gene co-expression network as a source of functional annotation for cattle genes

Area of Science:

• Genomics
• Bioinformatics
• Systems Biology

Background:

• Vertebrate genomes have a consistent number of protein-coding genes (~20,000).
• The cattle genome remains poorly annotated, with many genes lacking functional understanding.
• Gene co-expression analysis assumes similar expression patterns indicate related functions or pathways.

Purpose of the Study:

• To construct a genome-wide Cattle Gene Co-expression Network (CGCN).
• To identify gene modules and predict functions for unannotated genes.
• To provide a resource for cattle gene function research.

Main Methods:

• Utilized 72 microarray datasets (1470 arrays) from NCBI GEO and EBI ArrayExpress.
• Consolidated 16,607 probe sets representing 11,397 genes into 32 co-expression modules.
• Applied the 'guilt-by-association' principle for gene function prediction.

Main Results:

• Identified 32 co-expression modules enriched for Gene Ontology (GO) terms and Reactome pathways.
• Predicted functions for 132 previously unannotated genes.
• Discovered four novel hub genes with potential key regulatory roles in immune activation, RNA processing, and metabolism.

Conclusions:

• The CGCN offers valuable insights for experimental biologists in the cattle genome.
• Facilitates experiment design, result interpretation, and hypothesis development.
• The network is publicly available for broader research use.