Fig. 1

A Hierarchical clustering of the tools used for de novo repeat detection in P. pacificus based on 1-kb non-overlapping windows identified repeat finders with similar performance. The y-axis reflects the Euclidean distance between the binary vectors of the 1-kb windows. B The 20 most abundant patterns of common 1-kb windows reflected the clustering results. C The sum of nucleotides each repeat finder spanned in Mb was calculated and compared to the entire genome of P. pacificus. RED masked the most genomic regions with RepeatModeler2 following. D The comparison between each of the repeat finders and RepeatModeler2 showed that the level of agreement varies, ranging from 95.6% with RepeatMasker (C. elegans as a template) to 32% with Tantan. E The pie chart shows the amount of repetitive regions identified by RepeatModeler2. Class I TEs and specifically LINEs, LTRs and Penelope elements consist the majority of annotated TEs in the P. pacificus genome. F The pie chart shows the number of identified RepeatModeler2 regions for each type. The majority of annotated masked genomic regions is assigned to simple repeats