Another reason to conduct replication studies is to extend the generalizablity of the association. It is important to know if the association exists and has similar magnitude in different environmental or genetic backgrounds. It is particularly interesting to know how these associations play out in populations of non-European ancestry, considering most GWA studies to date have been conducted in European-ancestry samples. Differences in allele frequencies and local linkage disequilibrium (LD) patterns across populations present both challenges and opportunities for replication and fine mapping. On the one hand, a marker allele that is strongly associated with a trait in one population may not have a detectable association in another, as the allele frequency may be smaller or the LD with the (unknown) causal variant may be much weaker. Thus, initial replication studies should focus on populations with genetic ancestry similar to that sampled in the study that first observed the marker-trait association, using the exact strategy outlined in the next section. Once credible evidence for this association has been established, replication efforts in other populations should type not only the marker
