There are also practical implications related to finding the variants responsible for observed associations. Perhaps the most important of these is that targeted sequencing within a “block” of LD surrounding GWAS discoveries is often not expected to identify the causal sites. Because modest amounts of recombination can enhance synthetic associations, and because recombination must be exceptionally high to eliminate the possibility of genome-wide significant associations, one or more of the responsible causal sites could be a very considerable distance from the common variant showing a signal of association. This possibility is starkly illustrated by the sickle cell anemia example in which genome-wide significant synthetic associations span ∼2.5 Mb around the causal mutation, although heterosis may also influence this result. This possibility suggests that efforts to identify causal variants responsible for GWAS signals that concentrate on a region of high LD surrounding the implicated variant are not well motivated and are likely to miss many and perhaps most of any rare variants that contribute to synthetic associations (see, for example, [5]). The distance over which synthetic associations occur also offers an
