Sequence-based characterization of bacterial communities has long been a hostage of limitations of both 16S rRNA gene and whole metagenome sequencing. Neither approach is universally applicable, and the main efforts to resolve constraints have been devoted to improvement of computational prediction tools. Here, we present semi-targeted 16S rRNA sequencing (st16S-seq), a method designed for sequencing V1–V2 regions of the 16S rRNA gene along with the genomic locus upstream of the gene. By in silico analysis of 13 570 bacterial genome assemblies, we show that genome-linked 16S rRNA sequencing is superior to individual hypervariable regions or full-length gene sequences in terms of classification accuracy and identification of gene copy numbers. Using mock communities and soil samples we experimentally validate st16S-seq and benchmark it against the established microbial classification techniques. We show that st16S-seq delivers accurate estimation of 16S rRNA gene copy numbers, enables taxonomic resolution at the species level and closely approximates community structures obtainable by whole metagenome sequencing.

中文翻译：

---

通过将 16S rRNA 基因序列与相邻基因组背景连接，实现高分辨率微生物组分析

基于序列的细菌群落特征长期以来一直是 16S rRNA 基因和整个宏基因组测序的限制因素。这两种方法都不是普遍适用的，解决约束的主要努力都致力于改进计算预测工具。在这里，我们提出了半靶向 16S rRNA 测序 (st16S-seq)，这是一种设计用于对 16S rRNA 基因的 V1-V2 区域以及该基因上游的基因组位点进行测序的方法。通过在 silico通过对 13 570 个细菌基因组组装体的分析，我们表明基因组关联的 16S rRNA 测序在分类准确性和基因拷贝数识别方面优于单个高变区或全长基因序列。使用模拟群落和土壤样本，我们通过实验验证 st16S-seq 并将其与已建立的微生物分类技术进行对比。我们展示了 st16S-seq 提供了对 16S rRNA 基因拷贝数的准确估计，实现了物种水平的分类学分辨率，并非常接近通过全基因组测序获得的群落结构。