Fragmented landscapes of mobile transposable element families bolster the diversity of large-effect alleles in Arabidopsis

Abstract

Transposable elements (TEs) are powerful endogenous mutagens, but our understanding of the 'mobilome' - the set of mobile TEs at the species level - has largely been shaped by the study of a single reference genome. To overcome this reference-centric view, we produced long-read genome assemblies for 91 Arabidopsis thaliana strains and performed a reference-free pangenomic graph analysis of structural variants. Our analysis revealed the presence of twice as many TE families with recent mobilization activity as previously described. These >200 newly identified TE families are ancestral, with a third experimentally shown to be mobile and although they are often characterized by the presence of full-length copies in only a few rather than in most strains, they nonetheless contribute significantly to the generation of large-effect alleles. These findings uncover a much more fragmented and impactful mobilome than previously documented and highlight how pangenome approaches dramatically improve the study of genetic diversity.