The predictable network topology of evolutionary genomic constraint.

Large-scale comparative genomics studies offer valuable resources for understanding both functional and evolutionary rate constraints. It is suggested that constraint aligns with the topology of genomic networks, increasing towards the center, with intermediate nodes combining relaxed constraint with higher contributions to the phenotype due to pleiotropy. However, this pattern has yet to be demonstrated in vertebrates. This study shows that constraint intensifies towards the network's center in placental mammals. Genes with rate changes associated with emergence of hibernation cluster mostly towards intermediate positions, with higher constraint in faster-evolving genes, which is indicative of a "sweet spot" for adaptation. If this trend holds universally, network node metrics could predict high-constraint regions even in clades lacking empirical constraint data.