Database-Pipeline Interactions Shape Taxonomic Inference in 16S rRNA Bacterial Community Profiling

High-throughput sequencing has transformed microbiome study by enabling cost-efficient, high-resolution profiling of microbial communities using 16S rRNA gene amplicons. Despite its widespread adoption, taxonomic inference remains highly dependent on the choice of reference database and bioinformatic pipeline. To provide a systematic benchmark, we evaluated three widely used 16S rRNA reference databases, Greengenes2 (GG2, v2022.10), SILVA 138.1, and RDP19, across representative analytical workflows, including QIIME2 (Q2; full-length and V4 classifiers) and a USEARCH/VSEARCH-based pipeline (UV). 12 publicly available mock community datasets spanning 37 individual sequencing events and varying 16S regions (V3-V4 and V4 region) were assessed using complementary qualitative (sensitivity, specificity) and quantitative (accuracy and error) metrics at the phylum, genus, and species levels. At the phylum level, most database-pipeline combinations achieved high sensitivity and specificity, indicating robust coarse-level classification. Differences became more pronounced at the genus level, where database-pipeline compatibility substantially influenced both detection and abundance estimation. Relative to their legacy predecessors, the tested database releases generally improved stability and quantitative fidelity. Under the tested versions and settings, the combination of UV and GG2 showed reduced false-positive rates and lower abundance error. Species-level classification remained challenging across all configurations, reflecting intrinsic resolution limits of 16S amplicons, although selected pairings provided comparatively improved performance. Overall, taxonomic performance was strongly resolution-dependent and shaped by interactions among database structure, classifier training region, and algorithmic design. This study provides version-specific baseline expectations for 16S rRNA gene-based taxonomic inference and highlights the importance of careful database-pipeline alignment and transparent methodological reporting in microbiome research.