Back

A simple model reveals why complex evolutionary innovations follow predictable paths

Jiang, D.; Pennell, M.; Sallan, L.

2025-10-16 evolutionary biology
10.1101/2025.03.27.645690 bioRxiv
Show abstract

The origin of morphological characters with novel forms and functions is among the most fascinating phenomena in evolution, and understanding its underlying principles has been a fundamental goal of evolutionary biology. An essential means by which novelties evolve is the divergence of repeated body parts, known as serial homologs, into various forms, as in the case of tetrapod limbs and insect wings. However, the mechanisms underlying such processes are poorly understood, with systematic investigations hindered by the lack of a generalized model that links selective, genetic, and developmental mechanisms and the expected patterns of evolutionary dynamics. To fill this gap, we propose a generalizable model for the evolution of serial homologs that is founded on recent advances in developmental biology, where the development of each serial homolog is controlled via a hierarchical gene regulatory network consisting of genes that specify character identity and those that encode the specific phenotypic state. Under this model, we investigated two types of evolutionary novelties: divergence in character states between body parts with conserved identities and switching of character identity by turning on and off master regulators. Using population genetic simulations, we demonstrate how selection and developmental constraints interact to shape dynamics of phenotypic evolution and conditions under which each type of novelty is likely to evolve. Together, our results provide general insights into how novelties could evolve from serial homologs, and offer a modeling framework where the developmental evolution of a broad range of phenotypic characters can be studied.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.