Illumina reads correction: evaluation and improvements.

Area of Science:

• Genomics
• Bioinformatics

Background:

• Illumina whole-genome sequencing (WGS) generates vast amounts of data.
• Accurate variant calling (VC) and de novo assembly depend on high-quality sequencing reads.
• Existing read correction tools require evaluation for modern sequencers like Illumina NovaSeq.

Purpose of the Study:

• To extensively evaluate existing Illumina WGS read correctors.
• To assess the impact of read correction on variant calling and de novo assembly.
• To introduce an optimized version of the RECKONER read corrector.

Main Methods:

• Evaluation of multiple existing read correction algorithms.
• Benchmarking correction impact on variant calling and de novo assembly pipelines.
• Testing algorithm performance on Illumina NovaSeq reads with varying quality characteristics.
• Optimization of the RECKONER algorithm with a new correction strategy.

Main Results:

• Read correction demonstrated improved variant calling quality in specific cases.
• Most evaluated algorithms are capable of processing Illumina NovaSeq reads.
• The new RECKONER version corrects high-coverage human reads in under 2.5 hours.
• RECKONER handles indel and substitution errors using a novel oligomer-based verification.

Conclusions:

• Read correction is a valuable step for improving WGS data analysis, particularly for variant calling.
• Existing correction algorithms are largely compatible with newer Illumina sequencing technologies.
• The enhanced RECKONER tool offers efficient and accurate read correction for WGS data.