An important question is whether population stratification can confound exome sequencing studies, and if so, how to correct for stratification in this context. Although excess-of-rare-variant tests are fundamentally different than single-variant tests, the possibility of stratification still exists because different ancestries within a structured population sample (e.g., African and European ancestry in African Americans, or northern European and southern European ancestry in European Americans) may have different allele frequency spectra due to their different demographic histories. For example, in an exome sequencing study in African Americans in which disease cases have more African ancestry than controls, one expects to see an excess of rare variants in cases, because African chromosomes carry more rare variants92.
