Benchmarking transposable element annotation methods for creation of a streamlined, comprehensive pipeline

Area of Science:

• Genomics
• Bioinformatics
• Molecular Evolution

Background:

• Advanced sequencing and assembly enable high-quality de novo genome assembly, including repetitive regions.
• Transposable elements (TEs) are crucial genomic components, but their annotation is challenging due to diverse types and repetitive nature.
• Existing TE annotation methods lack systematic comparison, hindering comprehensive genome-wide analysis.

Purpose of the Study:

• To systematically benchmark existing transposable element (TE) annotation tools.
• To develop a comprehensive computational pipeline for generating non-redundant TE libraries and performing whole-genome TE annotation.
• To facilitate a deeper understanding of TE diversity and evolution in eukaryotic genomes.

Main Methods:

• Benchmarking of TE annotation programs using a curated rice TE library.
• Evaluation of methods for annotating Long Terminal Repeat (LTR) retrotransposons, Terminal Inverted Repeat (TIR) transposons, Miniature Inverted Transposable Elements (MITEs), and Helitrons.
• Development of the Extensive de-novo TE Annotator (EDTA) pipeline for non-redundant TE library generation and annotation of intact, fragmented, and nested elements.

Main Results:

• Established performance metrics (sensitivity, specificity, accuracy, precision, FDR, F1) for comparing TE annotation tools.
• Created EDTA, a robust pipeline that produces filtered, non-redundant TE libraries.
• Demonstrated EDTA's effectiveness across plant (rice, maize) and animal (Drosophila) species, handling complex genomic regions with nested TE insertions.

Conclusions:

• The developed benchmarking results and EDTA pipeline significantly advance TE annotation in eukaryotic genomes.
• Comprehensive TE annotations derived from EDTA will enhance the study of TE diversity and evolution.
• EDTA is an open-source, freely available tool to support genomic research.