In a comment in the journal Nature today, Stanford researcher Carlos Bustamante and others argue that there has been a large bias in favor of people of European descent in genomic sequencing and the Genome-Wide Association Studies (GWAS) that follow. The story was picked up by other news agencies, such as this article on Wired.
Bustamante points out two differences: the amount of sequencing done on Europeans is greater, and the number of GWAS are vastly more (a whopping 96% are done on groups of European ancestry). But the astounding latter figure is perhaps less troubling than you might think.
This is because many GWAS can be done on the same data set, looking for genetic associations with different diseases. This means that a small bias in the sequencing done to create usable data can lead to a very large bias in the number of association studies performed. The researchers doing the association studies are simply using the data that are available. The real target should be to ensure that sequencing is done of more diverse groups, and the association studies would follow.
There is another reason why much of the sequencing effort has focused on Europeans (and East Asians): association studies are more powerful when the population being studied is more uniform. Populations in East Asia and Europe have had relatively little mixture from outside populations compared to populations in the Middle East or India, and this reduced the chances that such admixture will create a false signal. The tremendous amount of genetic diversity in Africa and its complicated history of human movements also make it more challenging than studies in Europe.
Bustamante is right to call for more studies of a more diverse sampling, particularly focused on sequencing. Not only will this help those populations medically, but will have the added benefit of providing diverse data that can help us to understand recent human evolution as well.
