College of Plant Protection, Hebei Agricultural University, Baoding 071001, China.
Department of Entomology, China Agricultural University, Beijing 100193, China.
Int J Mol Sci. 2019 Mar 1;20(5):1072. doi: 10.3390/ijms20051072.
Neuropterida is a super order of Holometabola that consists of the orders Megaloptera (dobsonflies, fishflies, and alderflies), Neuroptera (lacewings) and Raphidioptera (snakeflies). Several proposed higher-level relationships within Neuropterida, such as the relationships between the orders or between the families, have been extensively debated. To further understand the evolutionary history of Neuropterida, we conducted phylogenomic analyses of all 13 published transcriptomes of the neuropterid species, as well as of a new transcriptome of the fishfly species of Liu and Yang, 2006 (Megaloptera: Corydalidae: Chauliodinae) that we sequenced. Our phylogenomic data matrix contained 1392 ortholog genes from 22 holometabolan species representing six families from Neuroptera, two families from Raphidioptera, and two families from Megaloptera as the ingroup taxa, and nine orders of Holometabola as outgroups. Phylogenetic reconstruction was performed using both concatenation and coalescent-based approaches under a site-homogeneous model as well as under a site-heterogeneous model. Surprisingly, analyses using the site-homogeneous model strongly supported a paraphyletic Neuroptera, with Coniopterygidae assigned as the sister group of all other Neuropterida. In contrast, analyses using the site-heterogeneous model recovered Neuroptera as monophyletic. The monophyly of Neuroptera was also recovered in concatenation and coalescent-based analyses using genes with stronger phylogenetic signals [i.e., higher average bootstrap support (ABS) values and higher relative tree certainty including all conflicting bipartitions (RTCA) values] under the site-homogeneous model. The present study illustrated how both data selection and model selection influence phylogenomic analyses of large-scale data matrices comprehensively.
脉翅目是一个完整的高级阶元，包括广翅目（泥蛉、鱼蛉和齿蛉）、脉翅目和蛇蛉目。脉翅目内的几个高阶关系，如目之间或科之间的关系，一直存在广泛争议。为了进一步了解脉翅目的进化历史，我们对所有 13 种脉翅目物种的已发表转录组以及我们测序的 2006 年 Liu 和 Yang 鱼类的新转录组（广翅目：齿蛉科：齿蛉亚科）进行了基因组分析。我们的基因组数据矩阵包含来自 22 种全变态昆虫物种的 1392 个直系同源基因，代表脉翅目两个科、蛇蛉目两个科以及广翅目两个科作为内群分类单元，以及 9 个全变态目作为外群分类单元。系统发育重建分别采用了基于合并和合并的方法，在同质性模型和异质性模型下进行。令人惊讶的是，同质性模型下的分析强烈支持脉翅目为并系群，其中长翅目科被分配为所有其他脉翅目动物的姐妹群。相比之下，异质性模型下的分析结果表明脉翅目为单系群。同质性模型下，基于具有更强系统发育信号的基因（即较高的平均引导支持值和包括所有冲突二分法的相对树置信度值）的合并和合并分析也恢复了脉翅目的单系性。本研究全面说明了数据选择和模型选择如何影响大规模数据矩阵的基因组分析。
