Plant Molecular and Bioinformatics Center, Biological Sciences, Northern Illinois University, 1425 W. Lincoln Hwy, DeKalb, IL 60115-2861, USA.
Department of Africa and Madagascar, Missouri Botanical Garden, St. Louis, MO 63110, USA.
Mol Phylogenet Evol. 2021 Jun;159:107110. doi: 10.1016/j.ympev.2021.107110. Epub 2021 Feb 17.
In this study, we analyzed 313 plastid genomes (plastomes) of Poaceae with a focus on expanding our current knowledge of relationships among the subfamily Pooideae, which represented over half the dataset (164 representatives). In total, 47 plastomes were sequenced and assembled for this study. This is the largest study of its kind to include plastome-level data, to not only increase sampling at both the taxonomic and molecular levels with the aim of resolving complex and reticulate relationships, but also to analyze the effects of alignment gaps in large-scale analyses, as well as explore divergences in the subfamily with an expanded set of 14 accepted grass fossils for more accurate calibrations and dating. Incorporating broad systematic assessments of Pooideae taxa conducted by authors within the last five years, we produced a robust phylogenomic reconstruction for the subfamily, which included all but two supergeneric taxa (Calothecinae and Duthieeae). We further explored how including alignment gaps in plastome analyses oftentimes can produce incorrect or misinterpretations of complex or reticulate relationships among taxa of Pooideae. This presented itself as consistently changing relationships at specific nodes for different stripping thresholds (percentage-based removal of gaps per alignment column). Our summary recommendation for large-scale genomic plastome datasets is to strip alignment columns of all gaps to increase pairwise identity and reduce errant signal from poly A/T bias. To do this we used the "mask alignment" tool in Geneious software. Finally, we determined an overall divergence age for Pooideae of roughly 84.8 Mya, which is in line with, but slightly older than most recent estimates.
在这项研究中，我们分析了 313 个禾本科植物的质体基因组（质体基因组），重点是扩展我们对 Pooideae 亚科关系的现有认识，该亚科代表了超过一半的数据集（164 个代表）。总共为这项研究测序和组装了 47 个质体基因组。这是同类研究中最大的一次，包括质体水平的数据，不仅在分类和分子水平上增加了采样，旨在解决复杂和网状关系，而且还分析了在大规模分析中对齐间隙的影响，以及探索亚科的分歧，用一套扩展的 14 个公认的草化石进行更准确的校准和约会。纳入作者在过去五年内对 Pooideae 分类群进行的广泛系统评估，我们为该亚科生成了一个强大的系统发育重建，其中包括除两个超科（Calothecinae 和 Duthieeae）之外的所有分类群。我们进一步探讨了在质体分析中包含对齐间隙如何经常导致对 Pooideae 分类群之间复杂或网状关系的不正确或误解。这表现为在不同的剥离阈值（基于每个对齐列的百分比去除间隙）下，特定节点处的关系始终发生变化。我们对大规模基因组质体数据集的总结建议是，去除所有间隙的对齐列，以增加成对同一性并减少来自 poly A/T 偏倚的错误信号。为此，我们使用了 Geneious 软件中的“mask alignment”工具。最后，我们确定了 Pooideae 的总体分化年龄约为 84.8 Mya，与最近的估计值相符，但略老一些。
