From contigs towards chromosomes: automatic improvement of long read assemblies (ILRA)

Area of Science:

• Genomics
• Bioinformatics

Background:

• Long-read sequencing technologies offer advancements for genome assembly but present challenges with repeats and homopolymer errors.
• Existing methods often result in numerous contigs and inaccuracies, hindering complete genome reconstruction.

Purpose of the Study:

• To introduce and evaluate the ILRA (Improved Long Read Assembly) pipeline for correcting errors in long-read genome assemblies.
• To enhance the contiguity and accuracy of genome sequences generated using long-read technologies.

Main Methods:

• The ILRA pipeline first reorders, renames, merges, circularizes, or filters contigs from long-read data.
• Illumina short reads are then utilized to correct homopolymer indel errors within the assemblies.
• The pipeline's performance was benchmarked on various species, including Homo sapiens, Trypanosoma brucei, Leptosphaeria spp., and Plasmodium falciparum.

Main Results:

• ILRA successfully improved the quality of long-read genome assemblies up to 1 Gbp.
• Correction of homopolymer tracts reduced the misannotation of genes as pseudogenes.
• The study identified that an iterative application of the pipeline may be necessary for further error correction.

Conclusions:

• The ILRA pipeline is an effective tool for enhancing the quality of long-read genome assemblies.
• This approach addresses key limitations of long-read sequencing, improving genome sequence accuracy and contiguity.
• The pipeline is publicly available, facilitating its adoption in genomic research.