Extended Data Fig. 3: Optimization of parameters for the characterization of hyper-divergent regions.

a,b, The total detected hyper-divergent regions in Mb (x-axis) and the percent overlap of long-read and short-read hyper-divergent classification (y-axis) are shown (Methods). Each point corresponds to one of the combination of threshold parameters for the variant count and coverage fraction of 1 kb bin to be classified as hyper-divergent. Each point is coloured by the variant count threshold (a) or the coverage fraction threshold (b). c, The relationship between the total size of hyper-divergent regions detected by the optimized short-read or long-read based approach is shown. Each point corresponds to one of the 15 long-read sequenced isotypes. Total sizes of hyper-divergent regions detected by the short-read based approach are shown on the x-axis, and total sizes of hyper-divergent regions detected by the long-read based approach are shown on the y-axis. d, The overlap between hyper-divergent regions defined by the optimized short-read based approach and long-read based approach is shown. Each point corresponds to one of the 15 long-read sequenced isotypes. Total sizes of hyper-divergent regions detected by either short-read or long-read based approach are shown on the x-axis, and the percentages of hyper-divergent regions detected by both approaches are shown on the y-axis.