There are caveats to GWAS. The most widely used platforms do not do a good job of covering rare variants (genetic markers with minor allele frequencies less than 1%). In addition, LD patterns are population specific. To address this issue, the HapMap project analyzed DNA from populations with African, Asian, and European ancestry. However, GWAS platforms differ in how well they cover genetic variation for different populations, so this must be considered when analyzing any particular gene (or the genome) because it can affect how good the coverage is in your sample depending on its racial composition. GWAS are also complicated by the fact that multiple testing corrections must be applied when conducting such an incredibly large number of tests (e.g., for 1 million markers across the genome), and there is differing opinion about how best to do that (van den Oord, 2007). This issue will be discussed further in section “Analyzing Genetic Data.”
