Whether the goal is to identify predictors of outcomes or to discover new biology underlying a trait of interest, the capability of GWAS to identify true genetic associations depends upon the overall quality of the data. Even simple statistical tests of association are compromised in the context of genome-wide SNP data that have not been properly cleaned, potentially leading to false-negatives and false-positive associations. Additionally, problems with the overall data quality will likely affect downstream analyses and studies beyond the initial GWAS. For example, the National Human Genome Research Institute (NHGRI) actively maintains an online catalog of GWAS results and associated publications [6], which stimulates downstream studies of replication and characterization in independent populations. Compromised data quality in the discovery phase may lead to false positive results that are carried forward into replication studies at great cost both in time and expense. Also, the National Institutes of Health (NIH) now mandates that secure, encrypted copies of primary GWAS data funded by NIH be made publicly available (with controlled access) for secondary analyses. These accessible datasets are maintained by the National
