Multi-threading the generation of Burrows-Wheeler Alignment.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing generates vast amounts of data, requiring efficient analysis.
• Mapping sequencing reads to a reference genome is a critical and time-consuming step.
• Existing tools like Burrows-Wheeler Alignment (BWA) lack full multi-threading support for alignment generation.

Purpose of the Study:

• To introduce BWA-MT, a novel multi-threaded tool designed to accelerate the alignment generation process.
• To improve the efficiency of genome sequencing data analysis.

Main Methods:

• Developed BWA-MT by enhancing the existing BWA tool with multi-thread support.
• Evaluated BWA-MT performance using a 24-core, 128 GB memory system.
• Utilized the hg19 human genome reference and varied read sizes (1-40 million spots) for workload.

Main Results:

• BWA-MT demonstrated a maximum performance improvement of 3.66 times compared to BWA.
• The tool generated Sequence Alignment/Map (SAM) files identical to those produced by BWA.
• Performance gains are dependent on available computing resources.

Conclusions:

• BWA-MT is a highly effective and fast alignment tool for genome sequencing data.
• The multi-threaded approach significantly reduces the time required for read mapping.
• BWA-MT offers a valuable solution for accelerating bioinformatics workflows.