Data from: Phylogenetic affiliation of SSU rRNA genes generated by massively parallel sequencing: new insights into the freshwater protist diversity

Recent advances in next-generation sequencing (NGS) technologies spur progress in determining the microbial diversity in various ecosystems by highlighting, for example, the rare biosphere. Currently, high-throughput pyrotag sequencing of PCR-amplified SSU rRNA gene regions is mainly used to characterize bacterial and archaeal communities, and rarely to characterize protist communities. In addition, although taxonomic assessment through phylogeny is considered as the most robust approach, similarity and probabilistic approaches remain the most commonly used for taxonomic affiliation. In a first part of this work, a tree-based method was compared with different approaches of taxonomic affiliation (BLAST and RDP) of 18S rRNA gene sequences and was shown to be the most accurate for near full-length sequences and for 400 bp amplicons, with the exception of amplicons covering the V5-V6 region. Secondly, the applicability of this method was tested by running a full scale test using an original pyrosequencing dataset of 18S rRNA genes of small lacustrine protists (0.2–5 µm) from eight freshwater ecosystems. Our results revealed that i) fewer than 5% of the operational taxonomic units (OTUs) identified through clustering and phylogenetic affiliation had been previously detected in lakes, based on comparison to sequence in public databases; ii) the sequencing depth provided by the NGS coupled with a phylogenetic approach allowed to shed light on clades of freshwater protists rarely or never detected with classical molecular ecology approaches; and iii) phylogenetic methods are more robust in describing the structuring of under-studied or highly divergent populations. More precisely, new putative clades belonging to Mamiellophyceae, Foraminifera, Dictyochophyceae and Euglenida were detected. Beyond the study of protists, these results illustrate that the tree-based approach for NGS based diversity characterization allows an in-depth description of microbial communities including taxonomic profiling, community structuring and the description of clades of any microorganisms (protists, Bacteria and Archaea).

下一代测序（next-generation sequencing, NGS）技术的最新进展，通过例如揭示稀有生物圈等途径，推动了各类生态系统中微生物多样性解析研究的发展。当前，针对经PCR扩增的小亚基核糖体RNA（SSU rRNA）基因区域开展的焦标签高通量测序，主要用于解析细菌与古菌群落，极少用于原生生物群落的表征。此外，尽管基于系统发育的分类学评估被认为是最可靠的方法，但相似性分析与概率分析方法仍是当前用于物种分类归属的最常用手段。

在本研究的第一部分，我们将一种基于系统发育树的方法，与18S rRNA基因序列分类归属的两类常用方法——BLAST（碱基局部比对搜索工具，Basic Local Alignment Search Tool）和RDP（核糖体数据库项目，Ribosomal Database Project）——进行了对比。结果显示，除覆盖V5-V6高变区的扩增子外，该方法在近乎全长序列及400bp扩增子的分类鉴定中均表现出最高的准确性。

其次，我们通过一项全规模测试验证了该方法的适用性：使用了一套源自8个淡水生态系统的小型湖生原生生物（0.2–5 μm）18S rRNA基因的原始焦磷酸测序数据集。我们的研究结果显示：其一，通过序列聚类与系统发育分类归属鉴定得到的操作分类单元（operational taxonomic unit, OTU）中，不足5%的类群可通过公共数据库序列比对在已发表的湖泊相关研究中被检测到；其二，结合系统发育分析方法的NGS测序深度，使得我们能够解析那些在经典分子生态学研究中极少甚至从未被检测到的淡水原生生物进化枝；其三，系统发育分析方法在解析研究不足或高度分化的种群结构时，表现出更强的稳健性。

更具体地说，我们检测到了隶属于平藻纲（Mamiellophyceae）、有孔虫门（Foraminifera）、网粘菌纲（Dictyochophyceae）与眼虫纲（Euglenida）的全新推定进化枝。除原生生物研究领域外，上述结果还证实，基于NGS的多样性表征工作中采用的系统发育树方法，可实现微生物群落的全方位深度解析，包括分类学谱分析、群落结构刻画以及各类微生物（原生生物、细菌与古菌）进化枝的描述。