Combined evidence annotation of transposable elements in genome sequences

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Transposable elements (TEs) are mobile genetic sequences comprising a substantial portion of metazoan genomes.
• Current TE annotation heavily relies on single computational tools like RepeatMasker, potentially limiting accuracy.
• Advances in gene annotation highlight the benefit of integrating multiple evidence sources for improved models.

Purpose of the Study:

• To develop a robust pipeline for transposable element (TE) annotation by integrating multiple computational methods.
• To enhance the accuracy and comprehensiveness of TE models in genome sequences.
• To provide a framework for manual curation and reliable TE annotation using tools like Apollo.

Main Methods:

• Developed a combined evidence-model TE annotation pipeline integrating homology-based and de novo identification methods.
• Utilized multiple tools including RepeatMasker, BLASTER, TBLASTX, BLASTN, RECON, and TE-HMM.
• Applied the pipeline to Drosophila melanogaster Release 4 genomic sequences for proof of principle.

Main Results:

• Successfully annotated transposable element (TE) models in Drosophila melanogaster, identifying 6,013 TEs compared to 1,572 previously.
• The euchromatic TE fraction in D. melanogaster was revised to 5.3% from 3.86%.
• Discovered numerous small, abundant TE fragments and nested TE insertions, with 8.6% of TEs inserted into others.

Conclusions:

• The integrated pipeline significantly improves the detection and annotation of transposable elements (TEs), including complex and nested elements.
• This approach elevates TE annotation quality to be comparable to gene models.
• The pipeline is adaptable for annotating TEs in other genomic regions and species.