PrinTE: A Forward Simulation Framework for Studying the Role of Transposable Elements in Genome Expansion and Contraction

Christopher W. Benson, Ting-Hsuan Chen, Tim Lu, Cecilia Deng, Shujun OubioRxiv 2025 · DOI

TL;DR

This paper introduces PrinTE, a novel forward-time simulation framework for studying transposable element (TE) dynamics in genomes. PrinTE allows for highly customizable simulations of TE insertions, deletions, and mutations, enabling researchers to explore complex evolutionary scenarios and infer TE dynamics from empirical data.

What Problem Does This Paper Address?

Genome expansion and contraction are known to be influenced by transposable elements (TEs), but the precise mechanisms and historical dynamics driving these changes have remained largely enigmatic. This is largely due to the difficulty in directly observing past evolutionary events. Existing simulation tools often lack the flexibility to model complex TE behaviors, such as nested insertions, sequence divergence, and the formation of solo LTRs, which are crucial for understanding genome evolution. This paper aims to bridge this gap by providing a powerful and flexible simulation tool that can capture these nuances.

Methods

The core of this work is the development of PrinTE, a forward-time genome simulation framework. Key features include:

  • Customizable TE Dynamics: PrinTE allows users to define parameters for TE insertion, deletion, mutation (including substitution models like JC69 and K2P), and the formation of solo LTRs.
  • Two Simulation Modes:
    • Fixed-Rate Method: Users specify constant rates for TE insertion and deletion.
  • Variable-Rate Method: Rates dynamically adjust based on genomic features like the number of intact TEs, TE prevalence, and gene disruption, aiming to mimic biological forces.
  • Genome Structure Modeling: PrinTE can simulate gene sequences (CDS) and model their interactions with TEs, allowing for the study of purifying selection and large-scale structural variants.
  • LTR-RT Dating: A novel k-mer-based approach, Kmer2LTR, is integrated to estimate sequence divergence between LTR regions of LTR retrotransposons, crucial for inferring TE activity over evolutionary time.
  • Burn-in Phase: PrinTE can generate initial genomes with specified TE content, size, and superfamily abundance, or use outputs from other simulators like TEgenomeSimulator.

Datasets

The study utilizes publicly available fungal genomes for analysis and simulation. Specifically:

  • Pucciniomycotina Fungal Genomes: 58 genomes were downloaded from NCBI and JGI MycoCosm to study genome expansion and contraction in rust fungi.
  • Arabidopsis thaliana Genome: Used for simulating genome evolution under the variable-rate method.
  • Simulated Genomes: PrinTE generates synthetic genomes for various analyses, and these are compared against real genomic data.

The paper also references existing TE simulation tools (SimulaTE, ReplicaTE, SLIM, TESD, etc.) for benchmarking and comparison.

Key Insights and Ideas

  • TE Divergence as a Historical Record: The distribution of TE sequence divergence directly reflects their historical insertion and deletion dynamics, making it a valuable proxy for inferring past TE activity.
  • Solo LTRs and Genome Obesity: Counterintuitively, a high solo:intact LTR ratio (indicating strong purging of LTR-RTs) was positively correlated with larger genome sizes in Pucciniomycotina. This suggests that illegitimate recombination alone might not be sufficient to prevent genome obesity and that other processes are at play.
  • Myrtle Rust Genome Evolution: Simulations suggest that Austropuccinia psidii (myrtle rust) likely experienced a prolonged period of genome expansion driven by LTR-RTs, followed by a shorter, potent period of contraction.
  • PrinTE's Utility for Hypothesis Testing: The framework's flexibility allows for testing specific evolutionary hypotheses about TE dynamics and their impact on genome evolution, such as the role of TE defense mechanisms.

Key Results

  • PrinTE Performance: PrinTE demonstrated efficient memory usage (<2GB) and scalable runtime. Its LTR-RT dating approach (Kmer2LTR) showed high accuracy, detecting mutations up to 30% divergence, outperforming other methods at higher divergences.
  • Simulation Fidelity: PrinTE simulated LTR-RTs with higher detection rates (77%) compared to ReplicaTE (15%) and SimulaTE (8%), indicating better reflection of biological features like target site duplications.
  • Genome Size Dynamics: Simulations showed that a deletion rate fourfold higher than insertion led to a 45% genome size decrease, while reversing this ratio caused rapid genome bloating.
  • Pucciniomycotina Analysis: The study revealed a wide range of genome sizes (13.6 Mb to 1018.1 Mb) and LTR sequence content (0% to 67.4%) within Pucciniomycotina. Microbotryomycetes, with compact genomes, were largely devoid of LTR sequences, suggesting immunity to LTR-RT invasion as a strategy for maintaining small genomes.
  • Species-Specific Dynamics: Simulations of C. ribicola, P. coronata f. sp. avenae, and A. psidii revealed distinct TE dynamics: C. ribicola showed rapid, balanced insertion/deletion; P. coronata f. sp. avenae had slow insertion/removal; and A. psidii exhibited a history of expansion followed by contraction.

Limitations and Future Work

  • TE Defense Mechanisms: While PrinTE models gene disruption and selection, it does not explicitly simulate lethal insertions. Future work could incorporate more complex TE defense mechanisms and epigenetic regulation.
  • Sequence Identity Distributions: The current model uses normal distributions for sequence identity. More complex patterns, which may go beyond simple distributions, could be incorporated.
  • Functional Genomics Context: Integrating more detailed functional genomic context, such as specific gene functions and epigenetic regulation, could enhance simulation realism.
  • Insertion Site Preferences: Future development could include more sophisticated modeling of TE insertion site preferences beyond simple euchromatic/heterochromatic distinctions.
  • Narrower Phylogenetic Levels: The study suggests that exploring TE dynamics at narrower phylogenetic levels within fungi could reveal further interesting variations.

Summary updated

Abstract

Abstract Genome expansion and contraction are reportedly driven by transposable element (TE) activity, but the underlying dynamics remain enigmatic due to a lack of historical records tracing these changes. Here, we present PrinTE for versatile, forward-time simulation of whole-genome sequences with highly customizable transposon dynamics. Through simulations, we confirm that the distribution of TE sequence divergence reflects their historical insertion and deletion dynamics, which can be used to infer TE dynamic parameters through PrinTE simulations. We analyzed the pangenome of Pucciniomycotina , a subdivision of fungi containing myrtle rust ( Austropuccinia psidii ), which drastically expanded its genome size to 1018 Mb. Our analyses reveal that the best strategy for controlling genome size is to avoid the invasion of LTR retrotransposons (LTR-RTs). While illegitimate recombination (IR) is considered the most effective counteraction of LTR-RT invasions leaving only solo LTR remnants, we observed a strong positive correlation between solo:intact LTR ratio (strength of LTR-RT removal) and genome size ( r = 0.65), and a near-linear correlation between solo LTR count and genome size ( r = 0.98). This result suggests that IR alone may not effectively prevent genome obesity. Through simulation of Pucciniomycotina genomes, we proposed that A. psidii might experience a prolonged period of genome expansion followed by a short, potent, and likely ongoing period of contraction. PrinTE is freely available at https://github.com/cwb14/PrinTE.git .
BibTeX
@misc{benson2025printe,
  author = {Christopher W. Benson and Ting-Hsuan Chen and Tim Lu and Cecilia Deng and Shujun Ou},
  title = {{PrinTE: A Forward Simulation Framework for Studying the Role of Transposable Elements in Genome Expansion and Contraction}},
  year = {2025},
  doi = {10.1101/2025.06.04.657780},
  howpublished = {bioRxiv},
}

Loading related content...

Figure