Ray Nicolas, Currat Mathias, Excoffier Laurent
Genetics and Biometry Lab, Department of Anthropology and Ecology, University of Geneva, Geneva, Switzerland.
Mol Biol Evol. 2003 Jan;20(1):76-86. doi: 10.1093/molbev/msg009.
We report here a simulation study examining the effect of a recent spatial expansion on the pattern of molecular diversity within a deme. We first simulate a range expansion in a virtual world consisting in a two-dimensional array of demes exchanging a given proportion of migrants (m) with their neighbors. The recorded demographic and migration histories are then used under a coalescent approach to generate the genetic diversity in a sample of genes. We find that the shape of the gene genealogies and the overall pattern of diversity within demes depend not only on the age of the expansion but also on the level of gene flow between neighboring demes, as measured by the product Nm, where N is the size of a deme. For small Nm values (< approximately 20 migrants sent outwards per generation), a substantial proportion of coalescent events occur early in the genealogy, whereas with larger levels of gene flow, most coalescent events occur around the time of the onset of the spatial expansion. Gene genealogies are star shaped, and mismatch distributions are unimodal after a range expansion for large Nm values. In contrast, gene genealogies present a mixture of both very short and very long branch lengths, and mismatch distributions are multimodal for small Nm values. It follows that statistics used in tests of selective neutrality like Tajima's D statistic or Fu's F(S) statistic will show very significant negative values after a spatial expansion only in demes with high Nm values. In the context of human evolution, this difference could explain very simply the fact that analyses of samples of mitochondrial DNA sequences reveal multimodal mismatch distributions in hunter-gatherers and unimodal distributions in post-Neolithic populations. Indeed, the current simulations show that a recent increase in deme size (resulting in a larger Nm value) is sufficient to prevent recent coalescent events and thus lead to unimodal mismatch distributions, even if deme sizes (and therefore Nm values) were previously much smaller. The fact that molecular diversity within deme is so dependent on recent levels of gene flow suggests that it should be possible to estimate Nm values from samples drawn from a single deme.
我们在此报告一项模拟研究，该研究考察了近期空间扩张对一个种群内分子多样性模式的影响。我们首先在一个虚拟世界中模拟范围扩张，该虚拟世界由一个二维的种群阵列组成，这些种群与其邻居交换一定比例（m）的迁移个体。然后，在合并方法下，利用记录的人口统计学和迁移历史来生成一个基因样本中的遗传多样性。我们发现，基因谱系的形状和种群内多样性的总体模式不仅取决于扩张的时间，还取决于相邻种群间的基因流动水平，该水平由乘积Nm衡量，其中N是一个种群的大小。对于较小的Nm值（每代向外迁移个体数小于约20个），相当一部分合并事件发生在基因谱系的早期，而随着基因流动水平的提高，大多数合并事件发生在空间扩张开始时。对于较大的Nm值，范围扩张后基因谱系呈星形，错配分布是单峰的。相比之下，对于较小的Nm值，基因谱系呈现出非常短和非常长的分支长度的混合，错配分布是多峰的。由此可见，在空间扩张后，像 Tajima's D 统计量或 Fu's F(S) 统计量这样用于选择性中性检验的统计量，只有在具有高Nm值的种群中才会显示出非常显著的负值。在人类进化的背景下，这种差异可以非常简单地解释以下事实：对线粒体DNA序列样本的分析显示，狩猎采集者中错配分布是多峰的，而新石器时代后的人群中是单峰分布。实际上，当前的模拟表明，即使种群大小（因此Nm值）以前要小得多，近期种群大小的增加（导致Nm值更大）足以防止近期的合并事件，从而导致单峰错配分布。种群内分子多样性如此依赖于近期的基因流动水平这一事实表明，应该有可能从单个种群中抽取的样本估计Nm值。
