Positive phenotypic selection inferred from phylogenies
Positive phenotypic selection inferred from phylogenies
复制标题
DOI：
10.1111/bij.12649
复制
发表时间：
2016-05-01
影响因子：
1.9
通讯作者：
Venditti, Chris
中科院分区：
文献类型：
--
作者：
Baker, Joanna;Meade, Andrew;Venditti, Chris
表型进化的速率在自然界中变化很大，这些速率往往反映了自然选择的强度。在这里，我们概述了一种方法，用于检测异常变化的表型进化的速度在整个基因。我们引入了一个简单的新的分支特异性度量（V）/（B），该度量将沿着分支观察到的表型变化分为两个分量：（1）可归因于背景速率（（B））的分量，以及（2）可归因于偏离背景速率（（V））的分量。当形态进化速率的变化所引起的预期变化量是背景速率的两倍时（（V）/（B）>2），我们认为这是正表型选择。我们将我们的方法应用于六个数据集，在每个数据集中找到多个正选择的实例。我们的研究结果支持了越来越多的赞赏，传统的渐进的表型进化的观点很少得到支持，与更多的情节观点取而代之。这使人们不再将表型进化视为一个简单的同质过程，并促进了从遗传角度与宏观进化解释的协调，为深入了解基因型和表型之间的联系铺平了道路。当遗传数据不可用或无法获得时，检测正选择的能力对现存物种来说是一个有吸引力的前景，但当应用于化石数据时，它可以揭示否则不可能的深时间自然选择模式。
Rates of phenotypic evolution vary widely in nature and these rates may often reflect the intensity of natural selection. Here we outline an approach for detecting exceptional shifts in the rate of phenotypic evolution across phylogenies. We introduce a simple new branch-specific metric (V)/(B) that divides observed phenotypic change along a branch into two components: (1) that attributable to the background rate ((B)), and (2) that attributable to departures from the background rate ((V)). Where the amount of expected change derived from variation in the rate of morphological evolution doubles that explained by to the background rate ((V)/(B)>2), we identify this as positive phenotypic selection. We apply our approach to six datasets, finding multiple instances of positive selection in each. Our results support the growing appreciation that the traditional gradual view of phenotypic evolution is rarely upheld, with a more episodic view taking its place. This moves focus away from viewing phenotypic evolution as a simple homogeneous process and facilitates reconciliation with macroevolutionary interpretations from a genetic perspective, paving the way to novel insights into the link between genotype and phenotype. The ability to detect positive selection when genetic data are unavailable or unobtainable represents an attractive prospect for extant species, but when applied to fossil data it can reveal patterns of natural selection in deep time that would otherwise be impossible.