Broadscale Ecological Patterns Are Robust to Use of Exact Sequence Variants versus Operational Taxonomic Units.
Broadscale Ecological Patterns Are Robust to Use of Exact Sequence Variants versus Operational Taxonomic Units.
复制标题
DOI：
10.1128/msphere.00148-18
复制
发表时间：
2018-07-18
期刊：
影响因子：
4.8
通讯作者：
Martiny JBH
中科院分区：
文献类型：
--
作者：
Glassman SI;Martiny JBH
由于测序技术的突破，微生物生态学家在过去十年中对微生物组的理解取得了非凡的进步。这些进展对从农业到人类健康的各个领域都有广泛的影响。由于数据库的限制，大多数微生物生态学研究使用基于DNA序列相似性的分组方法来近似分类。关于如何最好地划分和近似这种分类法，仍然存在广泛的争论。在这里，我们使用大量基于现场的数据集来检查细菌和真菌，并得出结论，生态结果没有重大差异。因此，看来标准微生物群落分析对分箱方法的细节并不过分敏感。最近的讨论集中在最好的方法来划定微生物类群，无论是基于精确序列变异（ESV）或传统的操作分类单位（OTU）的标记基因序列。我们试图测试分组方法（ESV与97% OTU）是否影响大型实地研究的生态结论。该数据集包括靶向所有细菌（16S rRNA）和真菌（内部转录间隔区[ITS]）的序列，在三个收集时间内，在非生物条件下显著不同的多种环境中。尽管微生物丰富度存在定量差异，但我们发现所有α和β多样性指标在两种方法分析的样品之间高度正相关（r > 0.90）。此外，社区组成的优势类群没有不同的方法。因此，统计推断几乎无法区分。此外，ESV仅适度增加真菌和细菌多样性的遗传分辨率（分别为OTU丰富度的1.3和2.1倍）。我们的结论是，对于宽尺度（例如，所有细菌或所有真菌）α和β多样性分析，ESV或OTU方法通常会显示类似的生态结果。因此，虽然有充分的理由使用ESV，但我们不必质疑基于OTU的结果的有效性。微生物生态学家在过去的十年中，由于测序技术的突破，我们对微生物组的理解取得了非凡的进步。这些进展对从农业到人类健康的各个领域都有广泛的影响。由于数据库的限制，大多数微生物生态学研究使用基于DNA序列相似性的分组方法来近似分类。关于如何最好地划分和近似这种分类法，仍然存在广泛的争论。在这里，我们研究了两种流行的方法，使用一个大型的基于现场的数据集检查细菌和真菌，并得出结论，有没有重大差异的生态结果。因此，看来标准微生物群落分析对分箱方法的细节并不过分敏感。
Microbial ecologists have made exceptional improvements in our understanding of microbiomes in the last decade due to breakthroughs in sequencing technologies. These advances have wide-ranging implications for fields ranging from agriculture to human health. Due to limitations in databases, the majority of microbial ecology studies use a binning approach to approximate taxonomy based on DNA sequence similarity. There remains extensive debate on the best way to bin and approximate this taxonomy. Here we examine two popular approaches using a large field-based data set examining both bacteria and fungi and conclude that there are not major differences in the ecological outcomes. Thus, it appears that standard microbial community analyses are not overly sensitive to the particulars of binning approaches. Recent discussion focuses on the best method for delineating microbial taxa, based on either exact sequence variants (ESVs) or traditional operational taxonomic units (OTUs) of marker gene sequences. We sought to test if the binning approach (ESVs versus 97% OTUs) affected the ecological conclusions of a large field study. The data set included sequences targeting all bacteria (16S rRNA) and fungi (internal transcribed spacer [ITS]), across multiple environments diverging markedly in abiotic conditions, over three collection times. Despite quantitative differences in microbial richness, we found that all α and β diversity metrics were highly positively correlated (r > 0.90) between samples analyzed with both approaches. Moreover, the community composition of the dominant taxa did not vary between approaches. Consequently, statistical inferences were nearly indistinguishable. Furthermore, ESVs only moderately increased the genetic resolution of fungal and bacterial diversity (1.3 and 2.1 times OTU richness, respectively). We conclude that for broadscale (e.g., all bacteria or all fungi) α and β diversity analyses, ESV or OTU methods will often reveal similar ecological results. Thus, while there are good reasons to employ ESVs, we need not question the validity of results based on OTUs. IMPORTANCE Microbial ecologists have made exceptional improvements in our understanding of microbiomes in the last decade due to breakthroughs in sequencing technologies. These advances have wide-ranging implications for fields ranging from agriculture to human health. Due to limitations in databases, the majority of microbial ecology studies use a binning approach to approximate taxonomy based on DNA sequence similarity. There remains extensive debate on the best way to bin and approximate this taxonomy. Here we examine two popular approaches using a large field-based data set examining both bacteria and fungi and conclude that there are not major differences in the ecological outcomes. Thus, it appears that standard microbial community analyses are not overly sensitive to the particulars of binning approaches.