Functional contexts of adipose and gluteal muscle tissue gene co-expression networks in the domestic horse.

The study delineates gene co-expression network structures in adipose and gluteal muscle tissues of domestic horses, using multiple approaches to elucidate known and potential novel functional roles within and between networks.

Study Design and Methodology

• The researchers used Camoco, a co-expression network tool, to depict the gene expression patterns in adipose and gluteal muscle tissues obtained from 83 domestic horses across 5 breeds.
• The responses of genes to environmental variation, which were indicated by gene expression profiles, were harnessed to build two separate but genotypically diverse gene co-expression networks.
• The key purpose was to pinpoint gene clusters that were significantly co-expressed in each tissue, and then compare these clusters between the two networks. This helped to quantify the degree of commonality between genes co-expressed in both networks, as well as isolate those found in just one network.

Quantifying Known and Novel Functions

• Through a supervised process called ontological enrichment, the researchers measured the known biological functions represented by each network. Ontologies, like GO and KEGG, were used to quantify the known biological functions in each tissue. A huge percentage of terms from these ontologies (40.3% from GO and 41% from KEGG) had co-expression in at least one tissue. Some terms had co-expression in both tissues, but a few demonstrated single tissue co-expression. This suggested a possibility of functional differentiation as indicated by the functional annotation.
• In parallel, an unsupervised technique was utilized. Unknown functions were delineated through Markov clustering. This process identified sets of unannotated genes exhibiting similar patterns of co-expression within a tissue. The gene sets of each tissue were compared, and the researchers identified clusters of genes that either segregate by tissue in their co-expression patterns or show high levels of co-expression in both tissues.
• To understand the nature of these clusters, these were integrated with GO and KEGG ontologies. This enabled differentiation between gene sets featuring previously curated annotations and those consisting of unannotated gene sets, indicating the potential for discovering novel biological functions.

Mapping the Transcriptional Landscape

• The integration of these various data sets permitted the researchers to map the transcriptional landscape of muscle and adipose tissues. This sets up a framework for understanding gene function that could potentially be generalized to other tissues in horses and other species.