LR_Gapcloser: a tiling path-based gap closer that uses long reads to complete genome assembly

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Genome assembly is crucial for understanding genetic information, but many assemblies remain incomplete with gaps.
• Third-generation sequencing (TGS) long reads offer potential for gap closure, yet current methods are computationally intensive.
• A need exists for fast and memory-efficient long-read-based gap-closure tools to achieve complete genomes.

Purpose of the Study:

• To develop LR_Gapcloser, a novel tool for rapid and efficient genome gap closure using long reads.
• To evaluate LR_Gapcloser's performance against existing state-of-the-art tools in terms of speed, memory usage, and accuracy.
• To assess the utility of LR_Gapcloser in improving genome contiguity for complex genomes and in hybrid assembly strategies.

Main Methods:

• Development of LR_Gapcloser, a tool utilizing raw and error-corrected long reads from TGS platforms for gap filling.
• Testing LR_Gapcloser on various gap types (de novo, repeat-derived, real) across different genomes, including human CHM1 and Triticum urartu.
• Evaluation of hybrid assembly strategies combining LR_Gapcloser with different assemblers to assess contiguity and correctness.

Main Results:

• LR_Gapcloser demonstrated superior performance by closing more gaps faster, with lower error rates, and significantly reduced memory usage compared to two existing tools.
• The tool effectively utilized raw reads for gap filling, outperforming error-corrected reads in certain scenarios.
• Significant improvements in contig N50 were observed: a 132-fold increase for the human CHM1 genome (143 kb to 19 Mb) and a 40% increase for Triticum urartu.
• A proposed hybrid assembly strategy incorporating LR_Gapcloser yielded a new human CHM1 genome assembly with a contig N50 > 28 Mb, surpassing previous non-reference assemblies.

Conclusions:

• LR_Gapcloser provides a fast and efficient solution for closing genome assembly gaps and enhancing contiguity.
• The tool is applicable to diverse gap types and complex genomes, utilizing raw long reads effectively.
• Hybrid assembly strategies employing LR_Gapcloser show promise for generating reference-grade genome assemblies.