Genome-wide scan of 29,141 African Americans finds no evidence of selection since admixture
Gaurav Bhatia, Arti Tandon, Melinda C. Aldrich, Christine B. Ambrosone, Christopher Amos, Elisa V. Bandera, Sonja I. Berndt, Leslie Bernstein, William J. Blot, Cathryn H. Bock, Neil Caporaso, Graham Casey, Sandra L. Deming, W. Ryan Diver, Susan M. Gapstur, Elizabeth M. Gillanders, Curtis C. Harris, Brian E. Henderson, Sue A. Ingles, William Isaacs, Esther M. John, Rick A. Kittles, Emma Larkin, Lorna H. McNeill, Robert C. Millikan, Adam Murphy, Christine Neslund-Dudas, Sarah Nyante, Michael F. Press, Jorge L. Rodriguez-Gil, Benjamin A. Rybicki, Ann G. Schwartz, Lisa B. Signorello, Margaret Spitz, Sara S. Strom, Margaret A. Tucker, John K. Wiencke, John S. Witte, Xifeng Wu, Yuko Yamamura, Krista A. Zanetti, Wei Zheng, Regina G. Ziegler, Stephen J. Chanock, Christopher A. Haiman, David Reich, Alkes L. Price
(Submitted on 10 Dec 2013)
We scanned through the genomes of 29,141 African Americans, searching for loci where the average proportion of African ancestry deviates significantly from the genome-wide average. We failed to find any genome-wide significant deviations, and conclude that any selection in African Americans since admixture is sufficiently weak that it falls below the threshold of our power to detect it using a large sample size. These results stand in contrast to the findings of a recent study of selection in African Americans. That study, which had 15 times fewer samples, reported six loci with significant deviations. We show that the discrepancy is likely due to insufficient correction for multiple hypothesis testing in the previous study. The same study reported 14 loci that showed greater population differentiation between African Americans and Nigerian Yoruba than would be expected in the absence of natural selection. Four such loci were previously shown to be genome-wide significant and likely to be affected by selection, but we show that most of the 10 additional loci are likely to be false positives. Additionally, the most parsimonious explanation for the loci that have significant evidence of unusual differentiation in frequency between Nigerians and Africans Americans is selection in Africa prior to their forced migration to the Americas.
Haldane's Sieve 
“Additionally, the most parsimonious explanation for the loci that have significant evidence of unusual differentiation in frequency between Nigerians and Africans Americans is selection in Africa prior to their forced migration to the Americas.”
Nice paper. Another possible reason for the loci that have significant evidence of unusual differentiation would be that African Americans, on average, come from different regions of West Africa, and certainly not Nigeria only. I do find it puzzling that so many studies, especially American studies, assume that all West Africans are homogeneous in place of West African origin and genetic prehistory.
Very nice paper, congratulations!
– it is not clear whether the authors removed centromeric, gaps, and low quality regions before calculating the genome-wide p-values. Sometimes these regions tend to have outlier values for Fst and other tests; thus, removing them before calculating the p-values can change the significance threshold, and possibly allow to identify some sweep that was not significant before.
– it would be nice to see a PCA of the samples, colored by the dataset. This would allow to see if there are great differences due to the sequencing technique used, or bias specific to a dataset.
– Africa is bigger than the USA, china, and half of Europe all together (see http://flowingdata.com/2010/10/18/true-size-of-africa/ ). These samples come from all around the USA, and it seems that there is no clear indication of the origin of each individual. Thus, this dataset merges individuals who are originary of regions very distant geographically (e.g. like merging europeans and asians in the same population), and this may reduce the power to detect sweeps.
Moreover, I think that it would be worth to try some other test for selection, before making such a strong claim on that there are no selective sweeps in all africans. Fst is highly sensitive to the demographic scenario, and given that here the dataset is composed by indivuals originary of a very large continent, there it may be some bias in the results.
Among the other tests to detect selection signals, maybe I would use the CLR or the XP-CLR, although there it may be some problems due to the different ascertainment bias in the samples. I would exclude all the tests based on linkage disequilibrium decay, because the dataset is composed by GWAS chips, in which most SNPs are tag SNPs (thus, not in linkage). Which other tests could be applied to such a large dataset?
To to clear, I think the claim is that they do not detect selection since admixture in African-Americans (i.e. in the last few hundred years), not any claim about selection in Africa.
I see it now, thank you! Yes, the paper is about African-Americans, not Africans.
In any case, then, it is a bit weird to consider all the African Americans as a single population. They come from an area (Africa) which is much bigger than where the samples are taken (USA). Moreover, they come from a continent where the genetic differentiation is greater than within any other continent. It is certainly true that they have admixed after being moved to Americas, but, what are the basis to consider them a single population?
“To be clear, I think the claim is that they do not detect selection since admixture in African-Americans (i.e. in the last few hundred years), not any claim about selection in Africa.”
I got that. My statement is more general. I’ve noticed an inclination in many papers to use Yorubans as the reference population for studies of African Americans. I think it is a bit too general, given that many, if not the majority of slaves came from West African countries that are further to the west.
Conclusions regarding African Americans of this paper would likely be the same if one was to use, for instance populations from Senegal, or populations from Burkina Faso. Still, African Americans deserve this test to be done and not to have the nuance of they’re population history oversimplified.
Pingback: Most viewed on Haldane’s Sieve: December 2013 | Haldane's Sieve