Botanical Institute, Cologne Biocenter, University of Cologne, Cologne, Germany.
PLoS One. 2012;7(8):e43652. doi: 10.1371/journal.pone.0043652. Epub 2012 Aug 24.
A DNA barcode is a preferrably short and highly variable region of DNA supposed to facilitate a rapid identification of species. In many protistan lineages, a lack of species-specific morphological characters hampers an identification of species by light or electron microscopy, and difficulties to perform mating experiments in laboratory cultures also do not allow for an identification of biological species. Thus, testing candidate barcode markers as well as establishment of accurately working species identification systems are more challenging than in multicellular organisms. In cryptic species complexes the performance of a potential barcode marker can not be monitored using morphological characters as a feedback, but an inappropriate choice of DNA region may result in artifactual species trees for several reasons. Therefore a priori knowledge of the systematics of a group is required. In addition to identification of known species, methods for an automatic delimitation of species with DNA barcodes have been proposed. The Cryptophyceae provide a mixture of systematically well characterized as well as badly characterized groups and are used in this study to test the suitability of some of the methods for protists. As species identification method the performance of blast in searches against badly to well-sampled reference databases has been tested with COI-5P and 5'-partial LSU rDNA (domains A to D of the nuclear LSU rRNA gene). In addition the performance of two different methods for automatic species delimitation, fixed thresholds of genetic divergence and the general mixed Yule-coalescent model (GMYC), have been examined. The study demonstrates some pitfalls of barcoding methods that have to be taken care of. Also a best-practice approach towards establishing a DNA barcode system in protists is proposed.
DNA 条码是一段较短且高度多变的 DNA 区域，旨在促进物种的快速鉴定。在许多原生动物谱系中，缺乏特定于物种的形态特征使得通过光学显微镜或电子显微镜进行物种鉴定变得困难，而在实验室培养物中进行交配实验的困难也不允许鉴定生物物种。因此，测试候选条码标记以及建立准确的物种识别系统比在多细胞生物中更具挑战性。在隐种复合体中，潜在条码标记的性能不能通过形态特征作为反馈来监测，但是由于多种原因，选择不适当的 DNA 区域可能会导致人为的物种树。因此，需要先验的系统发育知识。除了鉴定已知物种外，还提出了使用 DNA 条码自动划分物种的方法。隐藻门提供了一个系统上特征良好和特征不良的群体的混合物，并在本研究中用于测试一些适用于原生动物的方法的适用性。作为物种鉴定方法，已经使用 COI-5P 和 5'-部分 LSU rDNA（核 LSU rRNA 基因的 A 到 D 结构域）对 Blast 在对采样不良到采样良好的参考数据库的搜索中的性能进行了测试。此外，还检查了两种自动物种划分方法的性能，即遗传距离的固定阈值和通用混合 Yule 合并模型 (GMYC)。该研究展示了条形码方法存在的一些缺陷，需要加以注意。还提出了在原生动物中建立 DNA 条码系统的最佳实践方法。
