A Bayesian method for jointly estimating allele age and selection intensity
A Bayesian method for jointly estimating allele age and selection intensity
复制标题
DOI：
10.1017/s0016672307008944
复制
发表时间：
2008-02-01
影响因子：
1.5
通讯作者：
Slatkin, Montgomery
中科院分区：
文献类型：
--
作者：
Slatkin, Montgomery
联合估计影响等位基因的过去选择强度和等位基因年龄的问题在贝叶斯框架中得到了表述。等位基因年龄在给定频率下的先验分布是由现有的群体遗传学理论得到的。假设选择强度的先验分布反映了对新突变体的正选择更可能是弱的而不是强的。通过开发一种适用于低频等位基因的重要采样方法来说明一般方法。这种方法既可以用在已知紧密连接的标记位点的单倍型，也可以用在可以推断出连接的祖先染色体片段的长度时。以非洲G6PD的A -等位基因为例说明了该方法的应用。由于等位基因频率和重组的变化本质上都是随机的，因此任何方法所能达到的精度都是有限的。
The problem of jointly estimating the intensity of past selection affecting an allele and the allele's age is formulated in a Bayesian framework. The prior distribution of allele age given its frequency is obtained from existing population genetics theory. The prior distribution of selection intensity is assumed to reflect the fact that positive selection on a new mutant is more likely to be weak than strong. The general approach is illustrated by the development of an importance sampling method applicable to low-frequency alleles. This method can be used either when the haplotypes of closely linked marker loci are known or when the lengths of linked ancestral chromosomal segments can be inferred. The method is illustrated with an application to the A - allele of G6PD in Africa. Because changes in allele frequency and recombination are both intrinsically stochastic, there are limits to the accuracy achievable with any method.