Donnelly P, Tavaré S
Department of Statistics, University of Chicago, Illinois 60637, USA.
Annu Rev Genet. 1995;29:401-21. doi: 10.1146/annurev.ge.29.120195.002153.
Genealogical or coalescent methods have proved very useful in interpreting and understanding a wide range of population genetic data. Our aim is to illustrate some of the central ideas behind this approach. The primary focus is genealogy in neutral genetic models, for which the effects of demography can be separated from those of mutation. We describe the coalescent for panmictic populations of fixed size, and its extensions to incorporate various assumptions about variation in population size and nonrandom mating caused by geographical population subdivision. The effects of such genealogical structure on patterns and correlations in genetic data are discussed. An urn model is useful for simulating samples at loci with complex mutation mechanisms. We give two applications of the genealogical approach. The first concerns methods for estimating the mutation rate from infinitely-many-sites data, and the second relates to inference about recent common ancestors and population history.
系谱法或溯祖法已被证明在解释和理解广泛的群体遗传数据方面非常有用。我们的目的是阐述这种方法背后的一些核心思想。主要关注点是中性遗传模型中的系谱，对于该模型，人口统计学效应可以与突变效应区分开来。我们描述了固定大小随机交配群体的溯祖过程，以及其扩展形式，以纳入关于群体大小变化和由地理群体细分导致的非随机交配的各种假设。讨论了这种系谱结构对遗传数据模式和相关性的影响。一个瓮模型对于模拟具有复杂突变机制位点的样本很有用。我们给出了系谱法的两个应用。第一个涉及从无限多位点数据估计突变率的方法，第二个与关于近期共同祖先和群体历史的推断有关。
