SIXPAC (Search for Interactions is Probably Approximately Complete) is an efficient, scalable search algorithm that finds synergy between pairs of physically unlinked SNPs (genome-wide) in large case-control datasets.