Detection, Characterization, and Biological Analysis of Long Tandem Repeats Using Nanochannel Technology

Many plant and animal genomes are notoriously difficult to assemble because of the vast amount of long tandem repeats which can span several hundred kilobases to multiple megabases. Although the repeat motifs can be sequenced and the amount of repeat material can be approximated by conventional sequencing technologies, it is often difficult or impossible to assemble them into long contigs, so the exact locations and copy numbers of these repeats often remain elusive, especially when the unit length exceeds the read lengths. Without knowing the genomic context of these repeats or the amount of repeat material, it is difficult to attach any biological relevance to them. Thanks to the recent developments in nanochannel technology which allow us to image intact megabase-scale molecules of DNA, repeat regions can be easily spanned and properly placed in the genome assembly. Using Irys® technology and novel algorithms designed specifically to investigate long repeat arrays, we were able to find, quantitate, and characterize previously enigmatic repeat regions in several plant and animal genomes. The genomic positions of these repeats were found by aligning non-repetitive portions of repeat-containing molecules or consensus de novo assemblies to the reference genome. The genomic context then provides insight into the biological significance of the repeats.

We find that repeat arrays are often much longer than predicted (or not detected at all) by short-read sequencing; for example, in a worm genome we observe a tandem repeat array spanning at least 350 kb, composed of a 40 kb motif repeat unit, which is not visible by any conventional sequencing methods. We also see trends in both plants and animals that may be of interest from an evolutionary standpoint, such as a prominent 9 kb repeat that is found in many plant genomes but not in animals. Using Irys genome maps, we can scaffold data generated from other sequencing methods, leading to a more complete genome assembly that is much better than either method alone. Furthermore, integration with sequence data allows us to infer the biological function of the repeats. We demonstrate this strategy with an Aribidopsis dataset, in which we find an 8.8 kb tandem repeat that corresponds to the production of rRNA. Our investigation shows that Irys technology is an effective tool for both completing genome assemblies and elucidating the biological significance of previously problematic repeat arrays.