ebwt2InDel

ebwt2InDel detects single nucleotide polymorphisms (SNPs) and insertions/deletions (INDELs) from the Burrows-Wheeler Transform (BWT) of a FASTA read set in a reference-free, alignment-free framework.

Key Features:

• Reference-Free and Alignment-Free Framework: The method operates without a reference genome or sequence alignment, relying solely on the BWT of the FASTA read set.
• Detection of SNPs and INDELs: The approach identifies both SNPs and INDELs from read data.
• Algorithmic Improvements: Analysis is conducted using only the BWT, removing the need for Suffix and LCP arrays and reducing memory consumption by approximately one order of magnitude compared to methods that require those structures.
• Performance Enhancements: A parallelized implementation on a 24-core machine is reported to be an order of magnitude faster than the sequential implementation.
• High Sensitivity and Precision: On synthetic datasets (Human chromosome 1 at 30x coverage) the method detected up to 83% of SNPs and 72% of INDELs, compared with 71% of SNPs and 51% of INDELs for state-of-the-art de Bruijn graph-based tools.
• Scalability: The approach has been validated on whole-genome sequencing real-world datasets while maintaining high sensitivity.

Scientific Applications:

• Genomic Variant Analysis: Applicable to variant discovery in genomic studies, including population genetics and personalized medicine.
• Large-Scale Genomics Projects: Suitable for whole-genome sequencing analyses in large-scale projects due to reduced memory footprint and parallel performance.
• Comparative Genomics: The reference-free design enables comparative analyses without relying on a fixed reference genome.

Methodology:

Builds on the framework of Prezza et al. (AMB 2019) by using the BWT to detect variants, addressing the fixed k-mer order and information loss in repeated regions, extending the framework to INDEL detection, and applying algorithmic refinements that eliminate the need for Suffix and LCP arrays.