Accurate high throughput alignment via line sweep-based seed processing.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• The increasing volume of sequencing data necessitates the development of accurate and fast alignment tools.
• Existing aligners face challenges in efficiently processing diverse read lengths, including short (Illumina) and long reads (Pacific Biosciences, Oxford Nanopore).

Purpose of the Study:

• To present a universal, high-accuracy, and performant alignment approach for both short and long sequencing reads.
• To enable efficient discovery of insertions and deletions arising from structural variants.

Main Methods:

• A universal three-stage alignment scheme is proposed.
• Introduction of two line sweep-based techniques: "strip of consideration" and "seed harmonization", replacing traditional chaining.
• Refined seeding strategy utilizing the FM-index.

Main Results:

• The developed approach demonstrates high accuracy and speed for short and long read alignments.
• Comprehensive comparisons confirm superior performance against state-of-the-art aligners in terms of accuracy and runtime.
• The method effectively supports the identification of structural variants, including insertions and deletions.

Conclusions:

• The presented universal alignment scheme offers a robust solution for modern sequencing data challenges.
• The novel algorithmic techniques provide efficient and accurate read alignment and structural variant detection.
• This approach is suitable for a wide range of genomic applications requiring precise and rapid sequence analysis.