Species-level resolution for the vaginal microbiota with short amplicons

作者：Qing Wei, Yiya Shi, Rongdan Chen, Yin'ai Zou, Cancan Qi, Yingxuan Zhang, Zuyi Zhou, Shanshan Li, Yiping Hou, Hongwei Zhou, Muxuan Chen · 发表于：mSystems · 年份：2024 · DOI：10.1128/msystems.01039-23 · 被引用次数：12 · 研究领域：Reproductive tract infections research、Urinary Tract Infections Management、Pelvic floor disorders treatments

Specific bacterial species have been found to play important roles in human vagina. Achieving high species-level resolution is vital for analyzing vaginal microbiota data. However, contradictory conclusions were yielded from different methodological studies. More comprehensive evaluation is needed for determining an optimal pipeline for vaginal microbiota. Based on the sequences of vaginal bacterial species downloaded from NCBI, we conducted simulated amplification with various primer sets targeting different 16S regions as well as taxonomic classification on the amplicons applying different combinations of algorithms (BLAST+, VSEARCH, and Sklearn) and reference databases (Greengenes2, SILVA, and RDP). Vaginal swabs were collected from participants with different vaginal microecology to construct 16S full-length sequenced mock communities. Both computational and experimental amplifications were performed on the mock samples. Classification accuracy of each pipeline was determined. Microbial profiles were compared between the full-length and partial 16S sequencing samples. The optimal pipeline was further validated in a multicenter cohort against the PCR results of common STI pathogens. Pipeline V1-V3_Sklearn_Combined had the highest accuracy for classifying the amplicons generated from both the NCBI downloaded data (84.20% ± 2.39%) and the full-length sequencing data (95.65% ± 3.04%). Vaginal samples amplified and sequenced targeting the V1-V3 region but merely employing the ...