Entropy-based High Performance Computation of Boolean SNP-SNP Interactions Using GPUs

It is being increasingly accepted that traditional statistical Single Nucleotide Polymorphism (SNP) analysis of Genome-Wide Asso- ciation Studies (GWAS) reveals just a small part of the heritability in complex diseases. Study of SNPs interactions identify additional SNPs that contribute to disease but that do not reach genome-wide signicance or exhibit only epistatic eects. We have introduced a methodology for genome-wide screening of epistatic interactions which is feasible to be handled by state-of-art high performance computing technology. Unlike standard software (1), our method computes all boolean binary interac- tions between SNPs across the whole genome without assuming a par- ticular model of interaction. Our extensive search for epistasis comes at the expense of higher computational complexity, which we tackled using graphics processors (GPUs) to reduce the computational time from sev- eral months in a cluster of CPUs to 3-4 days on a multi-GPU platform (2). Here, we contribute with a new entropy-based function to evaluate the interaction between SNPs which does not compromise ndings about the most signicant SNP interactions, but is more than 4000 times lighter in terms of computational time when running on GPUs and provides more than 100x faster code than a CPU of similar cost. We deploy a number of optimization techniques to tune the implementation of this function using CUDA and show the way to enhance scalability on larger data sets.