Fec: a fast error correction method based on two-rounds overlapping and caching.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Third-generation sequencing technologies provide long reads essential for genome analysis.
• High error rates in these long reads necessitate computationally intensive error correction.
• Existing error correction methods often perform redundant base-level alignments.

Purpose of the Study:

• To develop a fast and efficient error correction tool for long sequencing reads.
• To reduce the computational cost associated with error correction in high-coverage datasets.
• To improve the speed of genome assembly pipelines by optimizing the error correction step.

Main Methods:

• Introduced Fec, a novel error correction tool utilizing two-round overlapping and a caching mechanism.
• Employed a large window size in the first round for rapid initial overlap detection and correction.
• Utilized a small window size in the second round to refine corrections for reads with insufficient overlaps, reusing cached alignment data to avoid redundant computations.

Main Results:

• Fec demonstrated significant speed-ups across multiple datasets.
• Achieved 1.24-38.56 times speed-up compared to MECAT, CANU, and MINICNS on PacBio datasets.
• Achieved 1.16-27.8 times speed-up compared to NECAT and CANU on nanopore datasets.

Conclusions:

• Fec offers a substantial performance improvement for long-read error correction.
• The caching strategy effectively reuses alignment information, reducing computational overhead.
• Fec can be integrated into assembly pipelines or used as a standalone tool for efficient genome analysis.