Abstract
Taxonomic classification alone fails to capture the ecological and functional diversity of vaginal microbiomes, particularly those dominated by Gardnerella species. Using the expanded VIRGO2 gene catalog, we developed the vaginal inference of subspecies and typing algorithm (VISTA), a novel ortholog-based framework that defined metagenomic subspecies and 25 metagenomic community state types (mgCSTs), including six distinct Gardnerella-dominated profiles. The mgCSTs exhibit marked differences in species composition, functional gene content, transcriptional activity, and host immune responses. These findings reveal that Gardnerella predominance does not uniformly equate to dysbiosis and underscore the importance of functional context in shaping host-microbiome interactions. VISTA provides scalable classifiers and an interactive application to support mechanistic studies of vaginal microbiome function and its implications for reproductive health.IMPORTANCEThe vaginal microbiome plays a central role in reproductive and gynecologic health, yet its functional diversity and ecological organization remain poorly understood. Traditional 16S rRNA approaches provide only a partial view of this complexity, overlooking the strain-level variation that often determines microbial behavior and host outcomes. By applying metagenomic sequencing and scalable computational modeling, we developed the vaginal inference of subspecies and typing algorithm, a framework that defines gene-based subspecies and community state types across diverse populations. These classifications reveal new insights into the genomic and ecological foundations of vaginal community structure and offer a standardized resource for comparative and translational microbiome research. This work establishes the foundation for functionally informed diagnostics and precision interventions targeting women's reproductive health.