PlinkGPU: A Framework for GPU Acceleration of Whole Genome Data Analysis

Jeffrey Poznanovic
School of Informatics, University of Edinburgh
University of Edinburgh, 2010


   title={PlinkGPU: A Framework for GPU Acceleration of Whole Genome Data Analysis},

   author={Poznanovic, J.},



Download Download (PDF)   View View   Source Source   



Genome-wide association studies (GWAS) are performed in order to detect the genetic variations associated with physical traits (e.g. diseases), and Plink is a popular software system for analyzing the data of GWAS. Due to the large datasets involved, the task of data processing can be very time-consuming. Although GPUs (graphics processing units) are not generally applicable to all computing tasks, they have the capability to significantly speedup many types of time-intensive applications due to the GPU’s massively parallel architecture. The practical goal of this project is to investigate whether GPU acceleration of a pipeline of Plink’s time-intensive kernels can substantially improve its overall runtime performance. A framework called PlinkGPU has been designed to fulfill this goal – it includes CUDA and OpenCL implementations for a specific set of Plink operations. Real-life datasets of increasing size are used to test the performance and scalability of both implementations. With the implemented analysis on the largest dataset, PlinkGPU runs 94 times faster than the original CPU-based version of Plink. Finally, this paper discusses whether GPUs are the most appropriate strategy for this type of data-intensive analysis.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2021 hgpu.org

All rights belong to the respective authors

Contact us: