Unveiling the Human Nasopharyngeal Microbiome Compendium: Systematic Characterization of Community Architecture and Function Through a Comprehensive Meta-Analysis
Song, K.; Brochu, H. N.; Bustos, M. L.; Zhang, Q.; Icenhour, C. R.; Letovsky, S. N.; Iyer, L.
Show abstract
The nasopharyngeal microbiome acts as a dynamic interface between the human body and environmental exposures, modulating immune responses and helping to maintain respiratory health. Despite its clinical significance, the nasopharyngeal microbiome remains poorly characterized with few resources available. A robust standardized framework for distinguishing healthy from dysbiotic communities is needed to advance nasopharyngeal microbiome research. We conducted a comprehensive meta-analysis that yielded 28 independent nasopharyngeal microbiome studies encompassing 7,790 uniformly reprocessed samples from public repositories. Further, we developed a newly introduced, three-stage large language model-assisted background decontamination pipeline that mitigates study-specific technical and environmental variation while preserving biological signal integrity. We identified six reproducible nasopharyngeal community state types (NPCSTs), each dominated by distinct genera: Moraxella (I), Corynebacterium-Dolosigranulum (II), Staphylococcus (III), Haemophilus (IV), mixed (V), and Streptococcus (VI). NPCSTs were key drivers of the sample diversity rather than disease status, requiring NPCST-stratified downstream analyses. Co-occurrence network analysis and functional pathway predictions revealed NPCST-specific microbial interactions and metabolic signatures. To bridge the translational gap from research to clinic, we developed the Nasopharyngeal Microbiome Health Index (NMHI), an independent continuous wellness score that achieves an AUC of 0.90 and 0.92 in internal and external validations, respectively. Unlike binary disease classifiers, the NMHI quantifies nasopharyngeal health along a spectrum, enabling risk stratification for respiratory infections, treatment monitoring, and dysbiosis identification in PCR-negative symptomatic patients. Our comprehensive framework establishes reproducible NPCSTs, validated decontamination protocols, and publicly available classification models, providing the analytical framework for advancing nasopharyngeal microbiome research toward clinical implementation. The NMHI exemplifies international best practices for microbiome diagnostics through rigorous external validation, standardized pipelines, and clinically interpretable wellness metrics. It further delivers an evidence-based framework for translating nasopharyngeal microbiome research into precision respiratory medicine, enabling future development of microbiome-based respiratory therapeutics. Graphic AbstractAnalysis of 7,790 nasopharyngeal microbiomes using our newly introduced LLM-human hybrid background decontamination pipeline uncovered six stable and reproducible nasopharyngeal community state types (NPCSTs) that constitute fundamental ecological states, explaining 52% of microbial variance compared to just 13% from technical factors. These NPCSTs, each with characteristic dominant genera, diversity profiles, microbial networks, and metabolic signatures, reveal why previous studies yielded conflicting results and demonstrate the necessity of community-aware interpretation. The Nasopharyngeal Microbiome Health Index (NMHI) bridges research to clinic with 0.92 AUC on external validation, translates these biological insights into clinical practice, enabling continuous wellness monitoring and dysbiosis detection beyond conventional diagnostics. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=200 SRC="FIGDIR/small/673381v2_ufig1.gif" ALT="Figure 1"> View larger version (105K): org.highwire.dtl.DTLVardef@7919f1org.highwire.dtl.DTLVardef@963b34org.highwire.dtl.DTLVardef@1090683org.highwire.dtl.DTLVardef@18aa4d7_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIWe identified six reproducible nasopharyngeal community state types (NPCSTs) that transcend study-specific variation, providing a standardized framework for classifying nasopharyngeal microbiome structures across diverse populations. C_LIO_LIOur newly introduced three-stage decontamination pipeline combining LLM-assisted screening with expert curation successfully eliminates environmental artifacts in low-biomass nasopharyngeal samples while preserving biological signal integrity. C_LIO_LIDiversity metrics vary across NPCSTs independent of disease status, while co-occurrence networks and functional profiles show NPCST-specific patterns, collectively requiring community-stratified analyses to prevent misleading health-disease interpretations. C_LIO_LIThe Nasopharyngeal Microbiome Health Index (NMHI) shifts respiratory assessment from binary pathogen detection to continuous microbiome wellness monitoring, identifying individuals with dysbiotic nasopharyngeal microbiomes. C_LI
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Genome-centric analysis of short and long read metagenomes reveals uncharacterized microbiome diversity in Southeast Asians 97%
- Mechanical ventilation affects respiratory microbiome of COVID-19 patients and its interactions with the host 97%
- Multi-omic integration of microbiome data for identifying disease-associated modules 96%
Similar papers in this journal
- MCSPACE: inferring microbiome spatiotemporal dynamics from high-throughput co-localization data 97%
- Deep learning reveals functional archetypes in the adult human gut microbiome that underlie interindividual variability and confound disease signals 97%
- Single Amplified Genome Catalog Reveals the Dynamics of Mobilome and Resistome in the Human Microbiome 96%
Similar papers in this journal
Similar papers in this journal
- A human gut metagenome-assembled genome catalogue spanning 41 countries supports genome-scale metabolic models 97%
- Processing-bias correction with DEBIAS-M improves cross-study generalization of microbiome-based prediction models 96%
- No evidence for a common blood microbiome based on a population study of 9,770 healthy humans 96%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.