Ohio State Researchers Launch PrinTE to Map Jumping Gene Evolution

Researchers at The Ohio State University released a computational tool named PrinTE on Nov. 2, 2026, to trace how transposable elements, or jumping genes, expand and contract genome sizes. Described in Nature Communications, the simulator addresses the genome size paradox by evolving ancestral genomes forward in time to measure insertion and deletion dynamics in plants and fungi.

The Genome Size Paradox in Plants and Fungi

The corn genome stands as a prominent example of this expansion, with up to 85% of its DNA consisting of jumping genes. Most living organisms house DNA snippets that copy and paste themselves inside a genome, driving up total DNA volume without adding obvious structural complexity. Senior study author Shujun Ou, assistant professor of molecular genetics at Ohio State, noted that transposon activity over millions or tens of millions of years generates dense insertions and deletions.

First author Christopher Benson, a postdoctoral scholar in Ou’s lab, explained the mechanism behind this growth:

These are big components of a genome that maybe don’t necessarily have a noticeable impact, but they exist. They’re functioning in the background in these ways that we’re still trying to understand. But one thing that they do is impact genome size evolution heavily because they often insert as a copy-and-paste. You have the original copy and the duplicated copy. So if you do this tons of times, you get a really enormous genome.

Is there any value of having a genome that’s bigger or smaller? Ou asked. It’s paradoxical because if genome size is important, the larger genome plants must have more actual function or a more complicated structure. But in observation, they don’t.

Tracking Evolutionary History With PrinTE

To solve how genomes reach these massive sizes, Ou and Benson designed PrinTE as an enhancement of an older program called TEGenomeSimulator. While the previous software captured only a single snapshot of transposon infection of a genome at a single point in time, the new framework starts from an inferred common ancestor and evolves forward to the modern-day genome.

The research team collaborated with the New Zealand-based creators of the original software. For their initial test, the team focused on fungal genomes, specifically myrtle rust, whose genome has expanded to 1 billion base pairs due to transposable element activity.

We’ll play the simulator forward to try to make the present-day fungus. Sometimes it takes thousands of simulations to find the one that best fits, but this allows us to test those hypotheses that we’ve typically been looking at but not addressing in full depth.

Because the simulation is covering millions of generations of evolution, the developers integrated machine learning safeguards. Beginning a simulation is like walking in the dark, Ou said, because we only see the outcome, but we don’t know what happened in the past. There can even be times when we’re walking in the wrong direction. So we use machine learning to guide us to walk in the right direction, or closer to the right direction.

Insertion Bursts and Host Defense Mechanisms in Myrtle Rust

Simulations of myrtle rust revealed that the insertion rate of jumping genes was 4 to 5 times higher than the deletion rate, driving the excessively large genome. However, those bursts were eventually followed by a drop in insertions, which is a sign of tightened control of TE activity.

The finding suggests inserted transposons can eventually be purged by the host to suppress their activity, much like an immune system mounting a delayed response to an infection. Ou emphasized that the most effective way to control the genome size is to not have any infection in the first place.

Financial support for the project came from Ohio State’s STEM Education Faculty Startup Award, a U.S. Department of Agriculture postdoctoral fellowship, the U.S. National Science Foundation, and the Bioeconomic Science Institute in New Zealand.