High-fidelity (repeat) consensus sequences from short reads using combined read clustering and assembly.

Area of Science:

• Genomics
• Bioinformatics

Background:

• Accurate genome assembly remains challenging, particularly for repetitive DNA sequences, despite advances in genomic data generation.
• Existing read clustering algorithms aid in repeat identification but deriving reliable repeat consensuses from unassembled genomes is difficult.

Purpose of the Study:

• To develop and validate a method for generating trustworthy and representative repeat consensuses from unassembled genomic data using short reads.
• To address the underrepresentation and misassembly of repetitive elements in genome assembly.

Main Methods:

• Integration of repeat identification and genome assembly techniques.
• Development of an automated workflow for consensus building from clustered short reads.
• Testing across diverse use cases including non-model genomes and specific repeat structures.

Main Results:

• The combined methods successfully generated robust and representative repeat consensuses across all tested use cases.
• High-fidelity repeat consensuses were validated against existing tools and long-read data, confirming representation of transposable elements.
• The approach demonstrated the feasibility of generating reliable repeat consensuses from short reads in an automatable manner.

Conclusions:

• The developed workflow facilitates more efficient and less manual repeat characterization and annotation.
• This method is expected to significantly benefit genome studies, especially for non-model organisms with limited genomic resources.