Abstract
AbstractRepetitive elements (REs) are integral to the composition, structure, and function of eukaryotic genomes. Yet, RE dynamics remain understudied in many taxonomic groups, preventing holistic understanding of how genomes and species evolve. Here, we investigated REs across 601 insect species (20 orders) to better understand the RE landscape of insects and to evaluate automated RE annotation methods in the era of biodiversity genomics. We identified wide variation in the types and frequency of REs across insect groups. We quantified associations between REs and protein-coding genes and found an elevated frequency of associations in insects with abundant long interspersed nuclear elements (LINEs). Sequencing technology impacts RE detection; ∼36% more REs could be identified in long-read versus short-read assemblies. Long terminal repeats (LTRs) showed markedly improved detection in long-read assemblies (162% more), while DNA transposons and LINEs showed less respective technology-related bias. We illustrate fundamental challenges to efficient study of REs in diverse groups, showing that in most insect lineages, 25–85% of repetitive sequences were “unclassified” compared to only ∼13% of unclassified repeats in Drosophila species. Our findings suggest this RE-annotation bottleneck, driven largely by uneven taxonomic representation in RE reference databases, is worsening. Although the diversity of available insect genomes has rapidly expanded, the rate of community contributions to RE databases (essential for RE annotation) has not kept pace, preventing high resolution study of REs in most groups. We highlight the tremendous opportunity and need for the field of biodiversity genomics to embrace REs and suggest collective steps for making progress towards this goal.
Publisher
Cold Spring Harbor Laboratory