Ultrafast SNP analysis using the Burrows-Wheeler transform of short-read data.

Area of Science:

• Genomics
• Bioinformatics

Background:

• Conventional sequence-variation analysis relies on mapping results that are often redundant and biased.
• A novel, efficient method for single-nucleotide polymorphism (SNP) analysis is needed.

Purpose of the Study:

• To propose a straightforward and faster approach for SNP analysis using the Burrows-Wheeler transform (BWT) of short-read data.
• To evaluate the efficiency and accuracy of the BWT-based SNP detection method.

Main Methods:

• Utilized the Burrows-Wheeler transform (BWT) for simultaneous processing of short-read sequence fragments.
• Incorporated fragment depth of coverage (FDC) as supplementary data for SNP analysis.
• Defined minimum length for uniqueness (MLU) and FDC on the reference genome to predict method exceptions.

Main Results:

• SNP detection using BWT was substantially faster than traditional mapping methods, completing in minutes for exome/transcriptome data and 20 minutes for genome data on standard hardware.
• The BWT method demonstrated high agreement with a state-of-the-art tool, with predictable exceptions related to read fragment usage and sequencing depth.
• BWT and FDC computation times were less than mapping times for sufficiently large datasets.

Conclusions:

• The BWT-based approach offers a computationally efficient and rapid method for SNP discovery from short-read sequencing data.
• This method provides a viable alternative to conventional mapping-based analyses, particularly for large-scale genomic studies.
• Understanding MLU and FDC is crucial for interpreting SNP detection accuracy and potential limitations.