Using gene topologies to partition sex chromosomes into regions with distinct histories.

Abstract

Sex chromosomes often evolve through successive recombination arrests, forming evolutionary strata. Accurately describing these strata, their boundaries and history is crucial for studying sex chromosome evolution, but this can be very challenging, particularly in young sex chromosomes where X-Y or Z-W divergence is low or in highly rearranged sex chromosomes where strata are divided into pieces. Standard approaches to distinguish recently evolved strata involve combining signals from dS, Fst and gene-by-gene phylogenies, all of which can be noisy. Here, we propose a new method that uses the proportion of allele topologies observed on sex-linked genes between at least a pair of sister species, obtained from reconstructed X and Y sequences aligned to the same reference, along with an outgroup to infer ancestral alleles, with flexibility in the reconstruction pipeline. We implement a change-point analysis to objectively identify transitions in these proportions. Our method both delimitates strata and infers whether their evolutionary history is shared or convergent between species. We illustrate and test this method on Silene and Artemia species. Our new method is reliable and robust, globally confirming previous findings known for some of these species. The method is also sensitive, refining those previous findings and identifying regions that were not previously reported. We finally illustrate how this approach can inform scenarios that retrace the history of sex chromosomes in these species. Overall, this approach presents a tool that can be widely used and that complement existing methods to study sex chromosomes.