Accelerating the quality measurement of DNA with GPUs

Abstract

High-throughput ADN sequencing technologies, also known as Next Generation Sequencing (NGS), have a huge impact in the development of bioinformatic tools. NGS generate massive amounts of data that support biomedical research, and assessing the quality of these data is a crucial task but also an important computational challenge. In the last ten years Graphic Processing Units (GPUs) have emerged as one of the major paradigms for solving complex problems using parallel computing techniques on commodity hardware. This paper presents Sodium, a high performance tool that efficiently evaluates the quality of NGS data using GPUs. Our experimental evaluation, performed on low-cost hardware, shows promising results. In particular, we observe acceleration values of up to 2 digits when our proposal is compared with FastQC, one of the most popular tools to evaluate the quality of NGS sequences.