Application of GPGPU for Acceleration of Short DNA Sequence Alignment in Unipro UGENE Project

Konstantin Okonechinkov, German Grekhov, Konstantin Stepanyuk, Mikhail Fursov
Unipro Center Of Information Technologies
Unipro Center Of Information Technologies, Technical report, 2010


   title={Application of GPGPU for Acceleration of Short DNA Sequence Alignment in Unipro UGENE Project},

   author={Okonechinkov, K. and Grekhov, G. and Stepanyuk, K. and Fursov, M.},



Download Download (PDF)   View View   Source Source   Source codes Source codes




A dramatic increase of available sequencing datasets has resulted in the need of fast sequence alignment methods. Plenty of novel methods were proposed to perform the fast alignment of NGS data and some of them appeared to be rather effective, however a relatively small number of existing alignment tools use Graphic Processing Units (GPUs) to speed up the alignment procedure. Unfortunately these tools are available only as source code packages and have very limited support for existing biological data formats. In this report we describe UGENE Genome Aligner, an efficient GPU-accelerated tool for fast short reads alignment. This tool is created as an extension to Unipro UGENE bioinformatics toolkit, a popular open-source software package for molecular biologists. The performance benchmarks of the Genome Aligner demonstrated a 2x speedup in comparison with existing tools. The relative performance of the GPU-accelerated Genome Aligner was measured by comparison with the CPU version of the algorithm. In result the GPU-optimized search procedure showed ~6x speedup over the CPU version. The GPU-accelerated Genome Aligner is fully integrated into UGENE framework and freely available for download. The integration of the Genome Aligner allows it to be easily included into any research pipeline provided by UGENE and makes it available "out-of-the-box" for enduser without any additional efforts.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2021 hgpu.org

All rights belong to the respective authors

Contact us: