Correlated stabilizing selection shapes the topology of gene regulatory networks.

This article investigates how correlated stabilizing selection influences the topology of gene regulatory networks (GRNs) by shaping mutational correlations (pleiotropy) among gene expressions. Using individual-based simulations, the authors compare three genetic architectures: a gene regulatory network model, a multilinear quantitative genetics model with epistasis and pleiotropy, and a gene pleiotropy (GP) model with additive effects but gene-specific mutational covariance. They find that while all models evolve mutational correlations aligned with the direction of correlated selection, GRNs exhibit distinct constraints whereby co-expression and pleiotropy are primarily determined by local regulatory motifs—especially direct regulatory interactions—and network distance between genes. The study suggests that correlated stabilizing selection can promote network topologies that bring co-selected genes into close regulatory proximity, thereby partially reflecting past selection patterns on gene expression, though the extent of this effect in large, realistic networks remains uncertain.