Fig. 4

Duplication and deletion mechanisms in the splicing model. a PCA based on the genetic distances among different direct repeats from the class I PRE1, class II PRE1, and PRE1k lineage. We extracted the sequences in PRE1k and class II PRE1 that were homologous to the two direct repeats in class I PRE1, and we used the genetic distance matrix among these homologous sequences for PCA. The PRE1k represented the homologous cluster before the direct repeats were formed, and the class II PRE1 represented the homologous cluster after removal of a 46-bp sequence from the ancestral class I PRE1 by splicing. b Splicing model of tandem PRE1 for partial duplications and deletions. The top panel shows the three basic models for the formation of three different combined PRE1 types. The bottom panel shows that two tandem quasi-dimeric PRE1 form a combined quasi-oligomeric PRE1 or a combined PRE1 with degenerate direct repeats. c Genetic distance of the three direct repeats (repeat A, repeat B, and repeat C) from 82 quasi-oligomeric PRE1 based on principal component analysis (PCA). The genetic distance matrix among the three direct repeats was used for PCA. Repeat B cluster was between the repeat A and C clusters that supported the splicing model for the formation of quasi-oligomeric PRE1. Solid geometric symbols highlight five quasi-oligomeric PRE1 that are the most similar A repeats. d Genetic distances between repeat A, repeat B, and repeat C sequences from different quasi-oligomeric PRE1. π, nucleotide diversity; r, Pearson’s correlation coefficient. T test, the P values of paired T tests of genetic distances among quasi-oligomeric PRE1 in three direct repeats