Human Diversity: Our Genes Tell Where We Live

I summarize the main conclusions of Laurent Excoffier's article [it references and interprets [7] heavily]

1. Clustering of individuals from diverse populations around the globe, without any a priori knowledge of their ancestry assigns them to five distinct continental clusters

2. All persons of say Sub-Saharan African ancestry are recognized as such. Thus, the five clusters are clearly separable. While small numbers of individuals are found to belong to more than one cluster, the clusters themselves are clear-cut.

3. Population substructure is not detected in West Eurasians (incl. Indians), or East Asians. The essential unity of the Caucasoid and Mongoloid races is affirmed. Population subtracture, in the form of clear subclusters is detected in Sub-Saharan Africans. Again, early anthropological insights into the clear divisions between Bantu Negroids, San Bushmen, and two Pygmy groups.

4. Excoffier attacks the race concept by claiming that it postulated that individuals belonging to the same race are significantly more similar to each other than they are to members of other races. But that is a strawman argument. Races are defined by noting the differences between humans. Clearly, humans of different races all have hands, feet, legs, eyes, hearts, sex organs, etc. The greater part of our human genetic makeup is shared among races because we are all human. But, part of our genetic makeup is distinct among races, because the races are divisions of Homo Sapiens Sapiens that have accumulated distinct patterns of genetic makeup to the effect that they are now clearly separable.

Human diversity: our genes tell where we live

Excoffier L., Curr Biol. 2003 Feb 18;13(4):R134-6.

Summary

"A detailed genetic analysis of more than a thousand human subjects clusters them into five groups corresponding to major geographical regions. This new study shows that self-reported ancestry is a good predictor of one's genetic make-up."

...

Figure 1. Geographical location of the 52 population samples studied by Rosenberg et al. [7].

The barriers numbered 1 to 4 correspond to the sequential partition of the sampled populations into genetic clusters.

...

"Also interesting is the observation that with two clusters, individuals found in populations from Africa, Europe, North Africa, the Middle East and the Indian sub-Continent (mainly Pakistani populations) present a large majority of their genes as coming from the same population, whereas genes from the other hypothetical population are at a majority in individuals from East-Asia, Oceania and the Americas. This first division of the world (Figure 1, barrier 1) is at odds with previous results where a first split has often been observed between sub-Saharan Africans and non-Africans [11, 12 and 13]."

"Assuming that three populations are present (K = 3) leads to a split of individuals found in sub-Saharan Africa from those found in Europe, North-Africa, the Middle East and Pakistan (Figure 1, barrier 2). With K = 4, a cluster of Asiatic and Oceanian individuals separates from Amerindians (Figure 1, barrier 3). With K = 5, an Oceanian cluster appears (Figure 1, barrier 4), and we are left with the pleasant picture of a world divided into genetic clusters that closely correspond to five geographic regions: sub-Saharan Africa, East Asia, Oceania, the Americas and the rest, comprising Europe, North Africa and West Asia. With K = 6, a new genetic cluster made up essentially of individuals from a single Pakistani population emerges, showing that with the invocation of further clusters, single pop-ulations with peculiar allele frequencies stand out, probably because of isolation and founder effects."

"It thus seems that these five groups do correspond to major subdivisions of the human population. Rosenberg et al. [7] then attempted to examine further the internal genetic structure of these subdivisions. Sub-Saharan Africa presents clear additional levels of subdivisions, in keeping with previous results [14]. Amerindian populations also present substantial subdivisions corresponding to the five sampled populations. The other regions present less clear subdivisions, in the sense that the recovered populations do not correspond to collections of individuals found at the same location, or that individuals have genes originating from several clusters."

...

"The main conclusion of this work is that there is a very good agreement between the geographic and genetic assignment of individuals. The five major genetic clusters do correspond to five geographic regions. In other words, sampled sub-Saharan Africans are `all' genetically sub-Saharan Africans and native Amerindians are `all' genetically Amerindians. It would be highly misleading to conclude that Rosenberg et al. [7] have just rediscovered five basic races. The concept of race indeed assumes that members of a race are much more similar to each other than they are from members of other races: this is not what is found here. On the contrary, this study estimates that, if you consider two genes from two individuals in the same geographic region, they are on average only 4% more similar than two genes drawn from individuals belonging to different regions."

[7] N.A. Rosenberg, J.K. Pritchard, J.L. Weber, H.M. Cann, K.K. Kidd, L.A. Zhivotovsky and M.W. Feldman, Genetic structure of human populations. Science 298 (2002), pp. 2381–2385.

Source

Posted by Dienekes at April 4, 2003 03:20 AM | PermaLink