Comparison of commercial DNA extraction kits for whole metagenome sequencing of human oral, vaginal, and rectal microbiome samples.
Comparison of commercial DNA extraction kits for whole metagenome sequencing of human oral, vaginal, and rectal microbiome samples.
复制标题
用于人类口腔、阴道和直肠微生物样本全宏基因组测序的商业 DNA 提取试剂盒的比较。
DOI：
10.1101/2023.02.01.526597
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Wylie,Dennis
中科院分区：
文献类型：
--
作者：
Wright,MichelleL;Podnar,Jessica;Longoria,KaylaD;Nguyen,TienC;Lim,Sungju;Garcia,Sarina;Wylie,Dennis
简介DNA提取和测序技术的进步是破译微生物组与人类健康和病理学相关的意义的基础。与其前身(即，基于扩增的方法)相比，全基因组鸟枪测序(WMS)在使用中越来越受欢迎。然而，与基于扩增子的方法一样，WMS也容易受到DNA提取方法的偏差，这可能会损害测序和后续发现的完整性。本研究的目的是评估四种商业DNA提取试剂盒之间的系统差异，这些试剂盒经常用于微生物组的WMS分析。方法口腔、阴道和直肠拭子一式四份，由医疗保健提供者从五名参与者中收集，并随机分配到四种DNA提取试剂盒中的一种。每个提取试剂盒还提取了两个提取空白和三个重复的模拟群落样本。对所有样品用NovaSeq 6000完成WMS。测序和微生物群落分析使用非度量多维标度和成分偏差分析。结果提取试剂盒对不同样本和人体部位的微生物分类群的读数百分比有不同的偏倚。PowerSoil Pro套件在接近模拟社区的预期比例方面表现最好。虽然HostZERO对革兰氏阴性细菌有偏见，但该试剂盒在提取真菌DNA方面优于其他试剂盒。在临床样本中，与其他试剂盒相比，HostZERO产生的分配给不同地点的智人的读数比例较小，分配给细菌分类群的读数比例较高。然而，HostZERO似乎对微生物群落的代表性有偏见，并按部位表现出最大的分散性，特别是对阴道和直肠样本。结论4种常用的DNA提取试剂盒在用于人体微生物组的WMS分析时存在系统差异。考虑到研究设计和数据解释中的这种差异，对于保障微生物组研究的完整性和结果的重复性至关重要。
Introduction Advancements in DNA extraction and sequencing technologies have been fundamental in deciphering the significance of the microbiome related to human health and pathology. Whole metagenome shotgun sequencing (WMS) is gaining popularity in use compared to its predecessor (i.e., amplicon-based approaches). However, like amplicon-based approaches, WMS is subject to bias from DNA extraction methods that can compromise the integrity of sequencing and subsequent findings. The purpose of this study was to evaluate systematic differences among four commercially available DNA extraction kits frequently used for WMS analysis of the microbiome. Methods Oral, vaginal, and rectal swabs were collected in replicates of four by a healthcare provider from five participants and randomized to one of four DNA extraction kits. Two extraction blanks and three replicate mock community samples were also extracted using each extraction kit. WMS was completed with NovaSeq 6000 for all samples. Sequencing and microbial communities were analyzed using nonmetric multidimensional scaling and compositional bias analysis. Results Extraction kits differentially biased the percentage of reads attributed to microbial taxa across samples and body sites. The PowerSoil Pro kit performed best in approximating expected proportions of mock communities. While HostZERO was biased against gram-negative bacteria, the kit outperformed other kits in extracting fungal DNA. In clinical samples, HostZERO yielded a smaller fraction of reads assigned to Homo sapiens across sites and had a higher fraction of reads assigned to bacterial taxa compared to other kits. However, HostZERO appears to bias representation of microbial communities and demonstrated the most dispersion by site, particularly for vaginal and rectal samples. Conclusions Systematic differences exist among four frequently referenced DNA extraction kits when used for WMS analysis of the human microbiome. Consideration of such differences in study design and data interpretation is imperative to safeguard the integrity of microbiome research and reproducibility of results.