Allotetraploid origin of O. kokonorica
We concluded that O. kokonorica was an allotetraploid (2n = 4x = 40) based on: (1) k-mer analysis, in which two peaks were present, with the smaller peak at the doubled multiplicity of the major peak, and the result estimated by Smudgeplot (Figure S1); (2) the high degree of collinearity among the 20 assembled pseudochromosomes (Figure 1B; Figure S5); and (3) the results of genome synteny and K s (synonymous substitutions per synonymous site) values between O. kokonoricaand C. songorica described below (Figure 1C).
To investigate the structure of the O. kokonorica subgenomes, we conducted a synteny analysis between subgenomes of C. songoricaand our O. kokonorica pseudochromosomes. We found that eachC. songorica chromosome corresponded with a pair of O. kokonorica pseudochromosomes, and vice versa (Figure 1C). Based on genome synteny and K s values, we divided the O. kokonoricapseudochromosomes into A and B subgenomes. We found that high collinearity among the two genomes by visualizing the synteny blocks. For subgenomes A of the two species, each pair of chromosomes corresponded very well, while for subgenomes B, two pairs of large interchromosomal rearrangements were revealed between chromosomes B2 and B5, and between B1 and B8. When using O. thomaeum , a diploid close relative to the two allotetraploid species, as a reference, we found no interchromosomal rearrangement events between O. kokonorica and O. thomaeum (Figure 1D), indicating that these large interchromosomal rearrangements occurred specifically in C. songorica .