Reply to 'Selective effects of heterozygous protein-truncating variants'.
Reply to 'Selective effects of heterozygous protein-truncating variants'.
复制标题
回复“杂合蛋白质截短变体的选择性效应”。
DOI：
10.1038/s41588-018-0301-y
复制
发表时间：
2019
期刊：
影响因子：
30.8
通讯作者：
Sunyaev,ShamilR
中科院分区：
文献类型：
--
作者：
Cassa,ChristopherA;Weghorn,Donate;Balick,DanielJ;Jordan,DanielM;Nusinow,David;Samocha,KaitlinE;O'Donnell-Luria,Anne;MacArthur,DanielG;Daly,MarkJ;Beier,DavidR;Sunyaev,ShamilR
Cassa等人。回复-感谢布莱恩·查尔斯沃斯和威廉·希尔对我们的工作感兴趣，并提出了一个重要的问题。在突变选择平衡的假设下，我们的研究估计了杂合蛋白截断变异体(PTV)的选择系数。虽然样本中每个基因的罕见有害PTV的平均数量可以很好地由确定性突变-选择平衡表达来近似，但Charlesworth和Hill询问假设变异不会因遗传漂移而改变是否公平。确定性情景要求方差等于期望值；换句话说，它假设方差是由抽样的影响而不是漂移决定的。查尔斯沃思和希尔提出了包括漂移在内的平衡模型的分析，该模型表明，与推断的平均杂合选择系数相对应的基因的方差可能被大幅夸大。005Het。他们还质疑这种膨胀是否导致杂合选择系数总体分布的方差降低，以及漂移的纳入是否影响了我们的基因特异性估计。我们用模拟来研究这些问题，这些模拟在真实的人口学历史和纯粹的数据驱动的方法下明确地模拟了遗传漂移。模拟和数据都表明，确定性近似适用于强选择下的基因，包括带有=的基因。0 05
Cassa et al. reply—We kindly thank Brian Charlesworth and William Hill for their interest in our work and for raising an important question1. Our study estimated selection coefficients for heterozygous protein truncating variants (PTVs), shet, under the assumption of mutationselection balance2. Although the mean number of rare deleterious PTVs per gene in a sample can be well approximated by the deterministic mutation-selectionbalance expression, Charlesworth and Hill inquire whether it is fair to assume that the variance is unaltered by genetic drift. The deterministic scenario dictates that the variance is equal to the expectation; in other words, it assumes that the variance is dominated by the effects of sampling rather than by drift. Charlesworth and Hill present an analysis of the equilibrium model including drift, which suggests that variance might be substantially inflated for genes corresponding to the inferred mean heterozygous selection coefficient,. 0 05 het. They also question whether this inflation leads to a decrease in the variance of the overall distribution of heterozygous selection coefficients and whether the inclusion of drift might have affected our gene-specific estimates. We investigated these issues with simulations that explicitly model genetic drift in the presence of a realistic demographic history and with purely datadriven approaches. Both the simulations and data suggest that the deterministic approximation is applicable for genes under strong selection, including genes with=. 0 05