Rapid Evolution of Enormous, Multichromosomal Genomes in Flowering Plant Mitochondria with Exceptionally High Mutation Rates
Figure 5
Size distribution of repetitive content by the number of repeat pairs (left column) and total repeat length (right column).
Both datasets are based on all repeat pairs identified with BLAST by searching each genome against itself. Note that this method is different than counting individual repeat copies, which cannot be unambiguously identified when repeats exist in numerous partially overlapping copies, as they do in these genomes. For example, a repeat with four copies would be associated with six unique repeat pairs. Because of the enormous number of multicopy, overlapping repeats in S. conica, the total length of repeat pairs exceeds the size of the genome even though more than half of it is single-copy. For these same reasons, the distribution of repeat lengths in this figure differs from the repeat coverage statistics reported in Table 1, which consider what fraction of the genome is covered by repeats but not the total number of repeat pairs. The reported 50% coverage threshold represents the median of the total repeat length distribution.