Predicting biogeographical ancestry in admixed individuals – values and limitations of using uniparental and autosomal markers

When microsatellite profiles generated from crime scene samples do not match a known person, or eye-witness information is unreliable, highly informative uniparental and autosomal markers can help unveil biogeographical ancestry. However, as genetic admixture is becoming increasingly common in cosmopolitan societies, concern arises with their accuracy and suitability when dealing with samples from admixed individuals. Here we assess the ability to detect biogeographical ancestry in 85 individuals from self-declared Asian and European admixed families using a set of uniparental (Y and mitochondrial DNA) and autosomal single nucleotide polymorphisms, specifically selected to distinguish between these two biogeographical ancestries. Haplogroups and autosomal genotypes were investigated using STRUCTURE to detect levels of admixture. All haplogroups were characteristic of self-declared populations of origin. Overall, the autosomal markers inferred biogeographical ancestry more accurately in admixed individuals, showing no significant differences between observed and expected contribution from each population studied according to level of admixture, although some outliers were observed. We suggest a panel of highly informative autosomal and uniparental markers should be employed to infer biogeographical ancestry of an individual to help detect admixed ancestries.