Suarez B K, Crouse J D, O'Rourke D H
Am J Phys Anthropol. 1985 Jul;67(3):217-32. doi: 10.1002/ajpa.1330670307.
Ten-level synthetic gene frequency maps derived from a principal component analysis of seven polymorphic loci are displayed for a large sample of North Amerindian populations. These maps are useful for assessing population affinities over broad geographical regions and perhaps, as others have argued, for inferring recent migrations. The influence of European admixture is investigated by deleting highly admixed populations and regenerating the maps. In broad outline the resultant geographic patterning, while appearing more homogeneous, preserves many features of the maps that include the highly admixed samples--especially with respect to the Eskimo/non-Eskimo dichotomy. Further, in an effort to evaluate how varying the number of display levels affects patterning as well as interpretation, the maps were replotted at 5 and 20 levels. The 5-level maps are found to accentuate differences between the full data set and the less admixed data set, while the 20-level maps tend to obscure these differences.
针对大量北美印第安人群样本,展示了通过对七个多态性位点进行主成分分析得出的十级合成基因频率图。这些图谱有助于评估广泛地理区域内的人群亲缘关系,或许如其他人所主张的,也有助于推断近期的迁徙情况。通过剔除高度混合的人群并重新生成图谱,研究了欧洲人混合血统的影响。总体而言,所得的地理格局虽然显得更加均匀,但保留了包含高度混合样本的图谱的许多特征——特别是在爱斯基摩人/非爱斯基摩人二分法方面。此外,为了评估展示层级数量的变化如何影响格局以及解读,这些图谱被重新绘制为五级和二十级。发现五级图谱突出了完整数据集与混合程度较低的数据集之间的差异,而二十级图谱往往会掩盖这些差异。