Tissue-driven hypothesis with Gene Ontology (GO) analysis.

Most of the genes are under selective pressure to maintain their expression levels in the tissues. In a recent study, we have proposed a "tissue-driven" hypothesis stating that the stabilizing constraints on gene expression levels can be partitioned among tissues; tissues differ in their tolerance to gene expression variances; and the constraints on expression divergence is correlated with the constraints on sequence divergence. Here we further tested the "tissue-driven" hypothesis by sub-grouping genes into Gene Ontology (GO) categories. We examined the distribution of tissue expression distance of genes in the major GO categories in the tissues. We also examined the correlation between tissue expression distances and tissue sequence distances or tissue duplicate distances in the major GO categories. Our results have shown that the tissues-specific stabilizing constraints are generally not dominated by particular GO categories. It is also shown that sub-grouping genes into GO categories increased the sensitivity for detecting potential positive factors in expression divergence in the tissues.