While any human genome can be seen as the mixture of its ancestors, here we focus on individuals such as African-Americans in the USA; Afro Caribbeans in the UK; Central and South Americans with European descent; Ethiopians and North Africans and others, who trace their ancestry from a recent admixture event (less than 100 generations ago)20 between two or more human populations separated by at least 1000 generations of independent genetic drift. Over these few generations, recombination events created a unique tiling of ancestry blocks for each recently admixed individual21,22. Importantly, even if the current PS models assume so, there is no evidence on whether any expressed phenotype of an admixed genome should be seen as the linear sum of ancestry-specific effect sizes calculated along the genome, considering its specific local ancestry composition. In addition, the genetic effect sizes reported for certain ancestries may be more accurate than others, due to the greater amount of studies available10,23. Thus, local PS accuracy may vary along the genome and among individuals, depending on the relative fraction and on the particular tiling of a given ancestry each person has.
