GPU-Accelerated Genomic Sequence Alignment for Bioinformatics

EasyChair Preprint 13744

Abstract

The exponential growth of genomic data has necessitated the development of efficient and scalable computational techniques for sequence alignment, a fundamental task in bioinformatics. Traditional CPU-based methods often struggle with the increasing volume and complexity of genomic sequences, leading to significant delays in data processing and analysis. This paper explores the utilization of Graphics Processing Units (GPUs) to accelerate genomic sequence alignment, leveraging their parallel processing capabilities to enhance performance and reduce computational time. We review current GPU-accelerated algorithms and frameworks, such as CUDA and OpenCL, highlighting their architecture and implementation strategies. Through a series of benchmark tests, we demonstrate the substantial improvements in alignment speed and efficiency achieved by GPU-based approaches compared to conventional CPU-based methods. Additionally, we discuss the challenges and limitations associated with GPU acceleration, including memory management and algorithm optimization, and propose potential solutions to address these issues. Our findings underscore the potential of GPU-accelerated genomic sequence alignment to transform bioinformatics workflows, enabling faster and more accurate analysis of large-scale genomic data, thereby facilitating advancements in personalized medicine, evolutionary biology, and other related fields.

Keyphrases: CPU-based methods, Graphics Processing Units (GPUs), OpenCL

@booklet{EasyChair:13744,
  author    = {Abill Robert},
  title     = {GPU-Accelerated Genomic Sequence Alignment for Bioinformatics},
  howpublished = {EasyChair Preprint 13744},
  year      = {EasyChair, 2024}}