Department of Plant Sciences, University of Oxford, Oxford, United Kingdom.
Mol Biol Evol. 2017 Dec 1;34(12):3267-3278. doi: 10.1093/molbev/msx259.
The correct interpretation of any phylogenetic tree is dependent on that tree being correctly rooted. We present STRIDE, a fast, effective, and outgroup-free method for identification of gene duplication events and species tree root inference in large-scale molecular phylogenetic analyses. STRIDE identifies sets of well-supported in-group gene duplication events from a set of unrooted gene trees, and analyses these events to infer a probability distribution over an unrooted species tree for the location of its root. We show that STRIDE correctly identifies the root of the species tree in multiple large-scale molecular phylogenetic data sets spanning a wide range of timescales and taxonomic groups. We demonstrate that the novel probability model implemented in STRIDE can accurately represent the ambiguity in species tree root assignment for data sets where information is limited. Furthermore, application of STRIDE to outgroup-free inference of the origin of the eukaryotic tree resulted in a root probability distribution that provides additional support for leading hypotheses for the origin of the eukaryotes.
任何系统发育树的正确解释都取决于该树是否正确定根。我们提出了 STRIDE，这是一种快速、有效且无需外群的方法，可用于在大规模分子系统发育分析中识别基因复制事件和种系树根推断。STRIDE 从一组无根基因树中识别出一组支持良好的内群基因复制事件，并对这些事件进行分析，以推断无根种系树的根位置的概率分布。我们表明 STRIDE 可以正确识别跨越广泛时间尺度和分类群的多个大规模分子系统发育数据集的种系树的根。我们证明，在信息有限的数据集中，STRIDE 中实现的新颖概率模型可以准确表示种系树根分配的歧义。此外，将 STRIDE 应用于真核生物树起源的无外群推断，得到的根概率分布为真核生物起源的主要假说提供了额外支持。
