Comparative phylogeographic inference with genome‐wide data from aggregated population pairs. (15th April 2020)

• Record Type: Journal Article
• Authors: Xue, Alexander T.; Hickerson, Michael J.
• Is Part Of: Evolution. Volume 74:Number 5(2020)
• Page Start: 808
• Page End: 830
• Publication Date: 2020-04-15
• Subjects: Aggregate site frequency spectrum (aSFS) -- approximate Bayesian computation (ABC) -- hierarchical co‐demographic modeling -- Multi‐DICE -- population genomic methods -- random forest; Evolution -- Periodicals; Heredity -- Periodicals; Évolution (Biologie) -- Périodiques; Hérédité -- Périodiques; 338.47004094
• DOI: 10.1111/evo.13945
• Languages: English
• ISSNs: 0014-3820

Abstract

Abstract: Comparing divergences across multiple sister population pairs has been a focus in phylogeography since its inception. Initial approaches used organelle genetic data and involved qualitative comparisons of phylogenetic patterns to evaluate hypotheses of shared and variable evolutionary responses. This endeavor has progressed with coalescent model‐based statistical techniques and advances in next‐generation sequencing, yet there remains a need for methods that can exploit aggregated genomic‐scale data within a unified analytical framework. To this end, we introduce the aggregate joint site frequency spectrum (ajSFS) by validating its use within a hierarchical Bayesian framework through several in silico experiments. Subsequently, we applied our method against two published restriction site–associated DNA marker datasets consisting of eight local replicates of a lamprey species pair and six co‐distributed passerine taxon pairs, respectively, with the aim of inferring variability in co‐divergence and co‐migration histories. We found that the lamprey population pairs exhibited temporal synchrony in both co‐divergence and collective secondary contact times, yet an idiosyncratic pattern in secondary migration intensities. In contrast, the bird population pairs displayed thoroughly asynchronous co‐divergence histories. Our results demonstrate that the ajSFS can be exploited for complex and flexible co‐demographic inference, opening up new possibilities for comparative phylogeography and population genomic studies.