Archive (2016–2006)

Cost-effective GPU-Grid for Genome-wide Epistasis Calculations

Journal: Methods of Information in Medicine
Subtitle: A journal stressing, for more than 50 years, the methodology and scientific fundamentals of organizing, representing and analyzing data, information and knowledge in biomedicine and health care
ISSN: 0026-1270

Focus Theme
Grid and Cloud Computing Methods in Biomedical Research
Guest Editors: M. J. A. Eugster, M. Schmid, H. Binder,
M. Schmidberger

Issue: 2013 (Vol. 52): Issue 1 2013
Pages: 91-95

Cost-effective GPU-Grid for Genome-wide Epistasis Calculations

Focus Theme - Grid and Cloud Computing Methods in Biomedical Research

B. Pütz (1), T. Kam-Thong (1), N. Karbalai (1), A. Altmann (1), B. Müller-Myhsok (1)

(1) MPI of Psychiatry, Statistical Genetics, Munich, Germany


Grid, Epistasis, GPU, computing


Background: Until recently, genotype studies were limited to the investigation of single SNP effects due to the computational burden incurred when studying pairwise interactions of SNPs. However, some genetic effects as simple as coloring (in plants and animals) cannot be ascribed to a single locus but only understood when epistasis is taken into account [1]. It is expected that such effects are also found in complex diseases where many genes contribute to the clinical outcome of affected individuals. Only recently have such problems become feasible computationally.

Objectives: The inherently parallel structure of the problem makes it a perfect candidate for massive parallelization on either grid or cloud architectures. Since we are also dealing with confidential patient data, we were not able to consider a cloud-based solution but had to find a way to process the data in-house and aimed to build a local GPU-based grid structure.

Methods: Sequential epistatsis calculations were ported to GPU using CUDA at various levels. Parallelization on the CPU was compared to corresponding GPU counterparts with regards to performance and cost.

Results: A cost-effective solution was created by combining custom-built nodes equipped with relatively inexpensive consumer-level graphics cards with highly parallel GPUs in a local grid. The GPU method outperforms current cluster-based systems on a price/performance criterion, as a single GPU shows speed performance comparable up to 200 CPU cores.

Conclusion: The outlined approach will work for problems that easily lend themselves to massive parallelization. Code for various tasks has been made available and ongoing development of tools will further ease the transition from sequential to parallel algorithms.

You may also be interested in...


J. Beyer 1 , C. Langer 1 , L. Fritz1 , M. Hadwiger 1 , S. Wolfsberger2 , K. Bühler1

Methods Inf Med 2007 46 3: 270-274

Maria Nastase Mannila1 , Per Eriksson1 , Carl-Göran Ericsson2 , Anders Hamsten 1,3, Angela Silveira 1

Thromb Haemost 2006 95 3: 420-427


K. Ichikawa (1), S. Date (2), T. Kaishima (1), S. Shimojo (3)

Methods Inf Med 2005 44 2: 265-269