Hidden in Plain Sight. How Ks histogram dynamics can reveal and obscure ancient whole genome duplications.

Abstract

Polyploidy is a significant force in angiosperm evolution, of great interest to evolutionary biologists and crop scientists. The mode of origin of a polyploid lineage is a continuum, ranging from whole genome duplication (WGD) from within the same species (autopolyploidy), to a WGD derived from hybridization between two (or more) distinct species (allopolyploidy). The polyploid mode of origin impacts many aspects of genome evolution, including patterns of chromosomal inheritance, homeologous exchange rate, and post-WGD diploidization. Here, we develop a novel polyploid genome simulation engine DemographiKS to demonstrate that the Ks histogram, one of the most broadly utilized methods for inferring WGD events, is highly sensitive to evolutionary parameters relating to the mode of origin, as well as demographic parameters such as migration events, population bottlenecks and expansions. Our results show that that the location of the Ks peak may correspond to the time of parental divergence (for allopolyploids); might be hidden at Ks=0 (for autopolyploids under a population contraction); may correspond to the mean time of coalescence of the diploid ancestor (for autopolyploids under a population expansion), or the time of most recent migration between the parental species (for allopolyploids). We also fit simulated Ks histograms to polyploid genomes from four well-studied plant lineages on the polyploid continuum (Coffea arabica, Zea mays, Populus trichocarpa, and Saccharum spontaneum), demonstrating that WGD simulations may be used to corroborate inferred demographies. In summary, we demonstrate that Ks histograms are information-rich, computationally tractable, and can be utilized to corroborate detailed evolutionary histories inferred by other methods.