Blumberg B S, Hesser J E
Proc Natl Acad Sci U S A. 1971 Oct;68(10):2554-8. doi: 10.1073/pnas.68.10.2554.
The hypothesis that loci differentially affected by selection can be identified was tested by a comparison of two American black populations. The hypothesis had previously been studied in Evans County, Georgia and the data from that study were compared with data obtained from the population of Sapelo Island, Georgia. In both, gene frequencies for 21 polymorphic systems were determined. Estimates were made for 14 of these traits of the gene migration (m) from whites into these black populations, by use of the equation: m = (q(n) - q(Af))/(q(w) - q(Af)), where q(w), q(n), and q(Af) are the frequencies of an allele in the white, black, and West African populations, respectively. In this calculation, it is assumed that the differences in gene frequencies between American blacks and Africans are due entirely to gene mixture. On this assumption, if only gene mixture were operating, the values of "m" determined for each locus would be about the same. If "m" has a different value at some loci than it has at others, then selection for traits controlled by genes segregating at these loci could account for this difference. Values of "m" in both populations were scattered. These values were ranked for each population and the rank orders compared. There was a highly significant association between the rank order of loci in the two populations. This finding supports the hypothesis that loci differentially affected by selection can be identified, and that the selective forces have operated on the same loci in both populations.
通过比较两个美国黑人种群，对“受选择差异影响的基因座可以被识别”这一假设进行了检验。此前在佐治亚州的埃文斯县对该假设进行过研究，并将该研究的数据与从佐治亚州萨佩洛岛种群获得的数据进行了比较。在这两个种群中，测定了21个多态系统的基因频率。利用公式：m = (q(n) - q(Af))/(q(w) - q(Af))，对其中14个性状估计了从白人到这些黑人种群的基因迁移率（m），其中q(w)、q(n)和q(Af)分别是白人、黑人及西非人群中等位基因的频率。在该计算中，假定美国黑人和非洲人之间的基因频率差异完全是由于基因混合造成的。基于这一假设，如果只有基因混合起作用，那么为每个基因座确定的“m”值将大致相同。如果某些基因座的“m”值与其他基因座不同，那么对这些基因座上分离的基因所控制性状的选择可能可以解释这种差异。两个种群中“m”值都呈离散分布。对每个种群的这些值进行排序，并比较排序顺序。两个种群中基因座的排序顺序之间存在高度显著的关联。这一发现支持了“受选择差异影响的基因座可以被识别”这一假设，以及“选择力在两个种群的相同基因座上起作用”这一观点。
