Transient ecological states shape inter-patient heterogeneity in the periodontal immune–microbial ecosystem

The oral cavity hosts a dynamic immune–microbial ecosystem in which interactions between host responses and microbial communities contribute to periodontal disease. However, how this biological mechanism is organized across patients, sites, and time remains poorly understood. Here, we asked whether longitudinal periodontal samples from the same patient maintain patient-associated immune–microbial similarity over time or undergo temporal and site-specific diversification. We addressed this question using distance-based sample trees to represent similarity among longitudinal cell, cytokine, and microbiome profiles from patients with comparable disease status followed for one year. We found that some patients exhibited patient-specific clustering, but these clusters did not consistently contain all longitudinal samples, implying patient-associated temporal and site-specific organization. Characterization of patient-specific clusters suggested candidate microbial and host-associated signatures, including patterns related to immune activity and tissue remodeling for some patients. Overall, the sample-tree framework provides a representation of longitudinal heterogeneity and a means of generating hypotheses about ecological states.