Efficient homology-based annotation of transposable elements using minimizers

Area of Science:

• Genomics
• Bioinformatics
• Plant Science

Background:

• Transposable elements (TEs) comprise over half of complex plant genomes and influence gene expression, impacting agronomically important traits.
• Long-read sequencing enables assembly of large plant genomes, but transposable element annotation remains a significant challenge.

Purpose of the Study:

• To introduce a novel functionality within the Next-Generation Sequencing Experience Platform (NGSEP) for efficient, homology-based transposable element annotation.
• To address the bottleneck in transposable element annotation for large and complex plant genomes.

Main Methods:

• Developed a new NGSEP feature that treats reference library sequences as long reads for mapping to genome assemblies.
• Implemented a hierarchical annotation assignment based on homology to the reference library.
• Evaluated the algorithm's performance on diverse plant species, including Arabidopsis thaliana, Oryza sativa, Coffea humblotiana, and Triticum aestivum.

Main Results:

• The NGSEP functionality demonstrated superior speed compared to traditional homology-based annotation tools, achieving speedups of 3x to over 20x.
• Reduced the annotation time for the bread wheat (Triticum aestivum) genome from months to hours.
• Successfully recovered up to 80% of TEs annotated by RepeatMasker with a precision of up to 0.95.

Conclusions:

• NGSEP provides a rapid and efficient solution for transposable element analysis, particularly in large and transposable element-rich plant genomes.
• This advancement facilitates more comprehensive genome annotation and characterization in plants.