(2/2)
Furthermore, it has been claimed that these clusters roughly match the five continents, and thus correspond to the five-race scheme for modern humans (sub-Saharan African, European, East Asian, Australoid, Amerindian). This was accomplished by the use of a computer algorithm called STRUCTURE. The problem with this analysis is that underlying assumptions are essentialist and do not correspond to the evolution- ary history of our species (Weiss & Long, 2009). A typical STRUCTURE-like analysis defined a number of parental populations (either by user input or statistical inference) and then assigns fractions of parental ancestry to individuals. Unrealistic assumptions are made about the nature of the parental populations, such as random mating, with Hardy–Weinberg equilibrium for allele frequencies, and minimal linkage disequilib- rium between parental populations. Of course, the most unsupported assumption of STRUCTURE-like analysis is that the world once harbored distinct and independently evolving populations (Weiss & Long, 2009). To determine the genetic character of the “parental populations,” one needs alleles that are “private” to the purported parental populations. However, these sorts of genetic markers, called ancestry informative markers were determined from small- to medium-size samples taken from geographically distant populations (Europe, West sub-Saharan Africa, and East Asia). Given this sampling, it is a circular exercise to structure individuals into clusters. Of course, if the AIMS are not really informative of population ancestry, a STRUCTURE analysis may be simply meaningless.
On balance, the modern population genetic analysis suggests that our species, anatomically modern humans, does not display biological races (Barbujani et al., 2013; Graves, 2005a, 2005b; Montagu, 1997). We do have geographically based genetic and physical variation, and this fact has been consistently misinterpreted as evidence of the existence of biological races in our species. However, our species is young, has very little genetic variation, has maintained relatively high levels of gene flow throughout its history, and thus has never reached the levels of differentiation between its sub- populations consistent with the identification of geographical races (see discussions of this in Templeton, 2002; also see Graves, 2011; Lawson Handley et al., 2007; Serre & Paabo, 2004). Unfortunately, most researchers involved in projects that are affected by genetic variation do not have training or understand the historical context of evolution- ary thinking concerning speciation and the role that geographical races play in the process (Nesse et al., 2010; White et al., 2009). This is illustrated by the fact that they often assume that socially defined races are legitimate biological races (Bliss, 2012; Friedman & Lee, 2013; Graves, 2010, 2011).
Watch this while you're at it.
youtube.com/watch?v=teyvcs2S4mI&t=1s