Hybrid CPU/GPU Acceleration of Detection of 2-SNP Epistatic Interactions in GWAS

Abstract

[Abstract] High-throughput genotyping technologies allow the collection of up to a few million genetic markers (such as SNPs) of an individual within a few minutes of time. Detecting epistasis, such as 2-SNP interactions, in Genome-Wide Association Studies is an important but time consuming operation since statistical computations have to be performed for each pair of measured markers. In this work we present EpistSearch, a parallelized tool that, following the log-linear model approach, uses a novel filter to determine the interactions between all SNP-pairs. Our tool is parallelized using a hybrid combination of Pthreads and CUDA in order to take advantage of CPU/GPU architectures. Experimental results with simulated and real datasets show that EpistSearch outperforms previous approaches, either using GPUs or only CPU cores. For instance, an exhaustive analysis of a real-world dataset with 500,000 SNPs and 5,000 individuals requires less than 42 minutes on a machine with 6 CPU cores and a GTX Titan GPU.