Global Landscape of Publicly Available Human Oral Microbiome Data.

Abstract

Despite rapid growth in oral microbiome research, it remains unclear how well publicly available data reflect the diversity of the global human population. This study systematically evaluated the geographic and sampling-type representativeness of publicly available human oral microbiome data. A global meta-research analysis of publicly available human oral microbiome records in the NCBI BioSample database released up to December 31, 2025, was conducted. Records were retrieved, harmonized, and analyzed across 4 dimensions: geographic origin, oral sampling type, temporal trends, and population-adjusted representation using a derived representation index (RI). A total of 222,454 BioSamples from 1,600 studies were identified, spanning 92 countries and 4 major oral sampling-type groups: oral fluids, oral mucosa and surfaces, dental plaque and calculus, and special or lesion-associated sites. Geographic distribution was highly concentrated; nearly half of all geographically annotated samples originated from the United States and China, while 61% of countries worldwide contributed no samples. Low- and middle-income regions, including Central and Southern Asia (RI = -12.76) and Sub-Saharan Africa (RI = -11.21), were underrepresented relative to their population sizes. Sampling-type distribution was similarly uneven, with saliva samples comprising more than half of all samples. In contrast, disease-relevant sites, including carious lesions, periapical lesions, and the dental pulp, each represented less than 0.2% of the dataset. Together, these findings underscore that publicly available human oral microbiome data remain unevenly distributed across geographic origin and sampling types, reflecting structural and practical factors that have persisted over time. Deliberate efforts to improve global representation, sampling diversity, and metadata standardization are needed to build a more scientifically robust oral microbiome evidence base.