Background:

Experimental proof of gene function assignments in plants is based on mutant analyses. T-DNA

insertion lines provided an invaluable resource of mutants and enabled systematic reverse genetics-based

investigation of the functions of Arabidopsis thaliana genes during the last decades.

Results:

We sequenced the genomes of 14 A. thaliana GABI-Kat T-DNA insertion lines, which eluded flanking

sequence tag-based attempts to characterize their insertion loci, with Oxford Nanopore Technologies (ONT) long

reads. Complex T-DNA insertions were resolved and 11 previously unknown T-DNA loci identified, resulting in about

2 T-DNA insertions per line and suggesting that this number was previously underestimated. T-DNA mutagenesis

caused fusions of chromosomes along with compensating translocations to keep the gene set complete

throughout meiosis. Also, an inverted duplication of 800 kbp was detected. About 10 % of GABI-Kat lines might be

affected by chromosomal rearrangements, some of which do not involve T-DNA. Local assembly of selected reads

was shown to be a computationally effective method to resolve the structure of T-DNA insertion loci. We

developed an automated workflow to support investigation of long read data from T-DNA insertion lines. All steps

from DNA extraction to assembly of T-DNA loci can be completed within days.

Conclusions:

Long read sequencing was demonstrated to be an effective way to resolve complex T-DNA insertions

and chromosome fusions. Many T-DNA insertions comprise not just a single T-DNA, but complex arrays of multiple

T-DNAs. It is becoming obvious that T-DNA insertion alleles must be characterized by exact identification of both

T-DNA::genome junctions to generate clear genotype-to-phenotype relations.