Peterson et al. 10 compare the commonly used self-reported racial/ethnic identity phenotype to an empirical approach that uses genetic data for assigning research subjects to genetically informed ancestry groups. Improved genetic analytics are needed as study samples become larger and more racially/ethnically inclusive. Given both recent and past migration patterns worldwide, ancestry can be difficult to determine, resulting in erroneous findings due to population stratification. Admixture is a well-known confounder in genetic studies, thus determining subjects’ population membership through self-report may no longer be sufficiently accurate for gene identification or association studies. This paper has a number of strengths including use of a large racially/ethnically diverse sample, use of a powerful empirical method (principal components analysis (PCA)) designed specifically for such classification purposes, and the use of genome-wide single nucleotide polymorphism (SNP) weights from well-defined reference panels from the 1000 Genomes Project. While finding a very high concordance (~95%) between self-reports and extended-PCA defined racial/ethnic groups, the findings indicate that empirical clustering methods provide an incremental increase in racial/ethnic group homogeneity and reduce marker loss and sample loss due to
