This project used insertion polymorphisms based on short interspersed elements (SINEs) as a potential alternative to SNPs as a means of inferring an individual's geographic ancestry or origin.
The inference of an individual's geographic ancestry can be critical in narrowing the field of potential suspects in a criminal investigation. Most current technologies accomplish this task by relying on single nucleotide polymorphism (SNP) genotypes. The disadvantage of this approach, however, is that SNPs can introduce homoplasy into an analysis, since they can be identical-by-state. On the other hand, SINE polymorphisms are identical-by-descent, essentially homoplasy-free, and inexpensive to genotype using a variety of approaches. The authors conducted a blind study using 100 Alu insertion polymorphisms to infer the geographic ancestry of 18 unknown individuals from a variety of geographic locations. Using a structure analysis of the Alu insertion polymorphism-based genotypes, the authors were able to infer correctly the geographic affiliation of all 18 unknown individuals with high levels of confidence. Twelve were identified as European, 1 as African/Afican-American, and 1 as Asian. The remaining four samples were classified as being of mixed ancestry; three were an admixture of European and African descent, and one was an admixture of Indian and Asian descent. Many of the probabilities of assignment were well over 80 percent, and the detection of admixture in individuals of mixed ancestry was easily accomplished. The Markov Chain Monte Carlo methodology used by the Structure 2.0 software package provided a powerful analysis to group all individual into the selected number of populations and then determine the probability that each individual belonged to any given group. In addition, the software can detect admixture between populations in individual genotypes going back several parental generations. 1 figure, 2 tables, and 40 references