Data from: Targeted gene enrichment and high-throughput sequencing for environmental biomonitoring: a case study using freshwater macroinvertebrates

Recent studies have advocated biomonitoring using DNA techniques. In this study, two high-throughput sequencing (HTS)-based methods were evaluated: amplicon metabarcoding of the cytochrome C oxidase subunit I (COI) mitochondrial gene and gene enrichment using MYbaits (targeting nine different genes including COI). The gene-enrichment method does not require PCR amplification and thus avoids biases associated with universal primers. Macroinvertebrate samples were collected from 12 New Zealand rivers. Macroinvertebrates were morphologically identified and enumerated, and their biomass determined. DNA was extracted from all macroinvertebrate samples and HTS undertaken using the illumina miseq platform. Macroinvertebrate communities were characterized from sequence data using either six genes (three of the original nine were not used) or just the COI gene in isolation. The gene-enrichment method (all genes) detected the highest number of taxa and obtained the strongest Spearman rank correlations between the number of sequence reads, abundance and biomass in 67% of the samples. Median detection rates across rare (<1% of the total abundance or biomass), moderately abundant (1–5%) and highly abundant (>5%) taxa were highest using the gene-enrichment method (all genes). Our data indicated primer biases occurred during amplicon metabarcoding with greater than 80% of sequence reads originating from one taxon in several samples. The accuracy and sensitivity of both HTS methods would be improved with more comprehensive reference sequence databases. The data from this study illustrate the challenges of using PCR amplification-based methods for biomonitoring and highlight the potential benefits of using approaches, such as gene enrichment, which circumvent the need for an initial PCR step.近年来的研究已倡导采用DNA技术开展生物监测。本研究评估了两种基于高通量测序（high-throughput sequencing, HTS）的方法：一是针对线粒体基因细胞色素C氧化酶亚基I（cytochrome C oxidase subunit I, COI）的扩增子条形码测序，二是使用MYbaits的基因富集技术（靶向包含COI在内的9种不同基因）。该基因富集方法无需进行聚合酶链式反应（polymerase chain reaction, PCR）扩增，因此规避了通用引物引发的扩增偏差。研究团队从新西兰的12条河流中采集了大型无脊椎动物（macroinvertebrate）样本，对其进行形态学鉴定、计数并测定生物量。从所有样本中提取DNA后，采用Illumina MiSeq测序平台开展高通量测序。研究人员分别以6种基因（其中3种未纳入最初选定的9种基因范围）或单独使用COI基因，对测序数据中的大型无脊椎动物群落进行表征分析。在67%的样本中，基因富集方法（使用全部9种基因）检测到的分类单元（taxon）数量最多，且在序列读长数量、丰度与生物量之间构建出最强的斯皮尔曼等级相关系数（Spearman rank correlations）。针对稀有（总丰度或生物量占比<1%）、中等丰度（1%~5%）及高丰度（>5%）的分类单元，其检出率的中位数均以基因富集方法（使用全部基因）为最高。本研究数据显示，扩增子条形码测序过程中存在引物偏差：在多个样本中，超过80%的序列读长仅来自单个分类单元。若拥有更全面的参考序列数据库，两种高通量测序方法的准确性与灵敏度均可得到提升。本研究的数据阐明了基于PCR扩增的生物监测方法所面临的诸多挑战，并凸显了诸如基因富集这类无需初始PCR步骤的技术方案所具备的潜在优势。