Such interactions (which are to our knowledge unknown) might differ from recognized Selleck GDC973 bacterial interactions in dental plaque or other selleckchem mineralized surfaces, such as in the spatiotemporal model of oral bacterial colonization [18]. Nonetheless, the partial correlation analysis (Additional file
2: Figure S3) revealed a number of positive correlations among certain genera (including Actinomyces, Fusobacterium, Porphyromonas, Prevotella, Streptococcus, and Veillonella) that agrees with recognized dental plaque interactions, and also with a recent study that demonstrated how key oral species interact in order to grow in concert on saliva [17]. Hence, there appear to exist tight linkages among distinct bacterial taxa across various ecological oral niches. Interestingly, the lack of
analogous positive correlations in apes suggests that other bacterial interactions may prevail in their oral cavity, which strengthens the overall distinctiveness of the Pan and Homo microbiomes. Conversely, there were also a number of positive correlations present in both humans and apes. Although the underlying reasons for those correlations remain check details unknown for now, they might indicate basic bacterial interactions that are robust across a variety of primate hosts. Our results provide only limited support for the concept of a taxon-based core microbiome, i.e. a set of microbial OTUs which are characteristic
of the saliva microbiome across a set of individuals/species, and hence may be important for the functional HSP90 requirements of the saliva microbiome. A previous study that found support for a core oral microbiome (~75% of the OTUs in the study) in healthy individuals [28] was based on just three individuals; the putative core microbiome that we identified for humans as well as for apes accounts for a much smaller fraction of the OTUs in our study (12.1% and 10.3% respectively), even though we only required core OTUs to be found in at least one individual from each group/species. Although it is possible that these putative core OTUs do exist in the other individuals but at too low a frequency to be detected, the depth of sequencing in this study was sufficient to detect (with 99% probability) on average any OTU present at a frequency of 0.9% or more. Thus, even if a core saliva microbiome does exist that was not detectable in the present study, it would seem to account for at most a small fraction of the OTUs that comprise the saliva microbiome. Alternatively, it may be that the core microbiome is defined functionally rather than taxonomically, such that different OTUs are able to provide the same functionality, as has been suggested for the gut microbiome [22, 32].