Dupanloup I, Bertorelle G
Dipartimento di Biologia, Università di Ferrara, Ferrara, Italy.
Mol Biol Evol. 2001 Apr;18(4):672-5. doi: 10.1093/oxfordjournals.molbev.a003847.
The relative contribution of two parental populations to a hybrid group (the admixture proportions) can be estimated using not only the frequencies of different alleles, but also the degree of molecular divergence between them. In this paper, we extend this possibility to the case of any number of parental populations. The newly derived multiparental estimator is tested by Monte Carlo simulations and by generating artificial hybrid groups by pooling mtDNA samples from human populations. The general properties (including the variance) of the two-parental estimator seem to be retained by the multiparental estimator. When mixed human populations are considered and hypervariable single-locus data are analyzed (mtDNA control region), errors in the estimated contributions appear reasonably low only when highly differentiated parental populations are involved. Finally, the method applied to the hybrid Canary Island population points to a much lower female contribution from Spain than has previously been estimated.
两个亲本群体对一个杂交群体的相对贡献（混合比例）不仅可以通过不同等位基因的频率来估计，还可以通过它们之间的分子差异程度来估计。在本文中，我们将这种可能性扩展到任意数量亲本群体的情况。新推导的多亲本估计器通过蒙特卡罗模拟以及通过合并人类群体的线粒体DNA样本生成人工杂交群体来进行测试。双亲本估计器的一般属性（包括方差）似乎被多亲本估计器保留了下来。当考虑混合的人类群体并分析高变单基因座数据（线粒体DNA控制区）时，只有当涉及高度分化的亲本群体时，估计贡献中的误差才显得相当低。最后，应用于加那利群岛杂交群体的方法表明，来自西班牙的女性贡献比之前估计的要低得多。
