Transposable elements, often referred to as “jumping genes,” have the remarkable ability to insert copies of their sequences into different locations within the genome.
With next-generation sequencing (NGS), obtaining the complete genomic sequence of an organism has become routine. This means we should also be able to track how transposable elements move within the genome using these sequence data.
However, NGS produces millions to billions of short reads, each only a few hundred bases long—like an enormous jigsaw puzzle. Reconstructing continuous chromosome sequences from these fragments requires extensive computation.
The presence of transposable elements complicates this process. When elements longer than the read length are scattered throughout the genome, assembly algorithms often generate multiple possible solutions, making it difficult to determine the correct one.
Long-read sequencing technologies can resolve this issue, but with short-read data alone, it remains challenging. Moreover, because identical copies of the target transposon are dispersed across the genome, simple similarity searches cannot reliably reveal where and how the element has moved.
After days of thinking about this problem, I realized that the target site duplication (TSD) created during transposition could serve as a key indicator.
Figure 1A shows the TSD associated with the Tos17 retrotransposon, which was my focus at the time. When Tos17 inserts, a 5-base duplication appears flanking its upstream and downstream ends.
By extracting short reads that contain the 5′ and 3′ ends of Tos17 and removing the Tos17 sequence itself, I collected the adjacent sequences (Figure 1B). When two sequences share the same 5-base motif at their ends, they form a pair—indicating the insertion site of Tos17.
If a reference genome is available, these paired sequences can be mapped directly to identify insertion sites. Even without a reference, the paired sequences themselves mark the location of Tos17 and can be used as markers.
The approach is simple: search short reads for transposon termini and pair the adjacent sequences. This can be implemented with a lightweight program.
I developed a prototype in Perl with fewer than 50 lines of code. Because it relies only on basic searches and pairing a small number of sequences, I believe this is one of the smallest and fastest programs for analyzing NGS data.
Feel free to download and try it:
https://github.com/akiomiyao/tif
Reference
Nakagome, M., Solovieva, E., Takahashi, A. et al. Transposon Insertion Finder (TIF): a novel program for detection of de novo transpositions of transposable elements. BMC Bioinformatics 15, 71 (2014). https://doi.org/10.1186/1471-2105-15-71