The nasal microbiome redefines Staphylococcus aureus colonisation

The nasal microbiome redefines Staphylococcus aureus colonisation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article The nasal microbiome redefines Staphylococcus aureus colonisation Dinesh Aggarwal, Katherine L. Bellis, Beth Blane, Marcus C. de Goffau, and 34 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6079410/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 02 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Staphylococcus aureus colonises the nose in humans, with individuals defined as persistent, intermittent or non-carriers. Unlike the gut microbiome, the nasal microbiome has not been studied in large numbers of people. Here, we define the nasal microbiome in ~ 1,100 individuals and combine this with S. aureus culture data. We identify seven community state types (CST), including two CSTs found more commonly in females. Approximately 70% of those who are persistently colonised with S. aureus have a single CST microbiome dominated by S. aureus , while non-carriers are distributed across the other six CSTs. Intermittent carriers are not a unique state but have microbiomes that resemble either non- or persistent carriers. Persistent carriage is positively associated with S. aureus abundance, and negatively associated with three Corynebacterium species, Dolosigranulum pigrum , Staphylococcus epidermidis , and Moraxella catarrhalis ; the microbiome can be exploited with machine learning to accurately predict S. aureus colonisation status. Finally, we find that certain S. aureus lineages are likely better adapted to colonisation. Our data provides a comprehensive view of the nasal microbiome with respect to S. aureus colonisation, describing two key states: a S. aureus dominated CST in which S. aureus shapes the microbiome, and a group of CSTs in which S. aureus is rare or absent. Biological sciences/Microbiology/Microbial communities/Microbiome Biological sciences/Microbiology/Pathogens Biological sciences/Microbiology/Microbial communities/Microbial ecology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction The human nose is populated by a range of bacterial species which constitute the nasal microbiota, including the important commensal and opportunistic pathogen Staphylococcus aureus 1 – 3 . S. aureus nasal carriage is clinically important; carriers are at greater risk of S. aureus infection, often caused by the colonising strain 4 – 6 whilst decolonisation can reduce infection rates 7 . Based on longitudinal sampling, S. aureus nasal colonisation states have historically been divided into persistent, intermittent, and non-carrier 8 . However, it has been hypothesised that there may be only two categories (persistent carriers and non-carriers); as compared to intermittent carriers persistent carriers: (a) have higher S. aureus loads 9 (b) are more likely to become recolonised and maintain colonisation in experimental colonisation, (c) will preferentially select their own strain when inoculated with multiple strains 6 , and (d) have higher antibody levels to a limited number of S. aureus antigens (in small-scale studies) 10 . Multiple host factors have been identified that influence S. aureus carriage, and colonisation is higher in adult males and children 11 – 13 , though many of these studies are small and in unrepresentative cohorts 12 – 14 . Interactions between the microbial residents of the anterior nares have been described between S. aureus and other members of the nasal microbiome 15 – 18 . For example, both Staphylococcus epidermidis 19 and Staphylococcus lugdunensis 20 have been shown to produce distinct compounds that inhibit S. aureus growth. Unlike the gut microbiome, only a limited number of small studies have investigated the nasal microbiome. Yan et al . examined the nasal microbiome of 12 individuals at the anterior nares and two sites in the inner part of the nasal cavity, revealing similarity in the dominant species between sites, but a lower overall diversity in the anterior nares 3 . Analysis of the microbiome composition of culture-defined persistent or non- S. aureus carriers revealed that S. aureus had both an antagonistic relationship with Corynebacterium pseudodiptheriticum and synergism with Corynebacterium accolens , that was confirmed experimentally 3 . A larger study of eighty-six older (mean age ~ 65) twin pairs defined seven community state types (CST; distinct groups of bacteria) in the nose, identified negative associations of S. aureus with Dolosigranulum , Simonsiella , Propionibacterium granulosum; a positive association with S. epidermidis , and found a lower overall bacterial density and S. aureus 16S gene copy abundance amongst women 17 . In early life, the assembly of the nasal microbiome is weakly influenced by the maternal microbiome, while environmental exposures such as daycare have a greater impact 18 , both the type of birth delivery and breast feeding influence the infant nasal microbiome with some variation between studies (reviewed in 21 , 22 ). It is clear that S. aureus colonises the nasopharynx in the first weeks of life 23 , and several species have been reported to support S. aureus colonisation in infants 18 , and be inversely correlated with maternal Dolosigranulum pigrum 8 . More recent metanalysis of paediatric (n = 99 individuals) and adult (n = 210) 16s rRNA gene-based studies revealed positive integrations between D. pigrum and both C. pseudodiptheriticum , and Moraxella nonliquefaciens in children 16 . This trinity of species have been reported to be associated with greater stability of the nasal microbiome in early life 24 . Similarly, in adults, four different Corynebacterium species have been found positively associated with D. pigrum , which in turn was negatively associated with S. aureus , an association that was further demonstrated in vitro 16 . The use of antagonistic bacterial strains as live biotherapeutics (probiotics) is an attractive option to reduce S. aureus nasal colonisation without the need for antibiotics. This concept was demonstrated using a Corynebacterium sp. to successfully eradicate S. aureus 25 . In more recent work, S. aureus was reported to be excluded from the gut in the presence of B. subtilis via inhibition of pathogen signalling 26 , a finding that was translated into a clinical trial of B. subtilis as a live biotherapeutic, which was successful eliminating viable S. aureus from the gut and reducing but not eradicating the bacterial loads in the nose 27 . In summary, while interactions between different bacterial species in the nasal microbiota have been identified including with S. aureus , nasal microbiome studies have only involved small sample sizes often in selected populations which likely reduce the generalisability of the findings. In addition, in recent years the importance of systematic removal of contamination in microbiome studies, particularly lower biomass/complexity environment such the nasal microbiome, has been established 28 – 30 . This means studies (particularly those with small sample sizes) that have not systematically removed contamination risk being confounded. Critically, for understanding the nasal microbiome in relation to S. aureus colonisation, no microbiome study greater than forty individuals 31 has yet included S. aureus colonisation status as defined by longitudinal sampling and culture, which has been used to understand S. aureus colonisation for 70 years 32 . Here, we utilise microbiome data from nasal swabs of 1,180 generally healthy community participants from across England in the CARRIAGE study, along with three weekly nasal swabs cultured for S. aureus to determine the microbiome structure associated with nasal S. aureus carriage, including evaluation of the validity of the current defined S. aureus colonisation states (persistent, intermittent and non-carriers). Results Determination of the nasal microbiome in a large cohort To study the biological basis of S. aureus colonisation we sampled generally healthy adult blood donors from across England with three self-taken nasal swabs taken at weekly intervals (Fig. 1 a). S. aureus colonisation status was assessed by culture, and was defined as: (i) persistent colonisation (306/1091 (28.0%); three S. aureus positive weekly nasal swabs, (ii) intermittent colonisation 191/1091 (17.5%); one or two positive swabs, and (iii) non-carrier 594/1091 (54.4%): no positive swabs, based on previous studies 9 , 10 , 33 , 34 (89 failed to return all samples). Participants had a mean age of 51.4 (median, 53) and 52.8% were female. A total of 1,756 samples, which included the first swabs of 1,180 participants underwent 16S rRNA gene sequencing to determine the microbiome composition ( Supplementary Fig. S1 ). After QC (see Methods and Supplementary Table S2 ), 1,055 samples remained, and after rarefaction and a systematic analysis to remove any likely contaminants, 53 Operational Taxonomic Units (OTUs) (24 species level taxa) remained. Differences in within-sample microbiome diversity We first investigated variation in Alpha diversity (measures of within-sample diversity) by S. aureus culture-defined S. aureus colonisation status, to determine differences in the microbiome between S. aureus colonisation states. We found Alpha diversity was significantly lower in samples from persistent carriers when compared to non-carriers or intermittent carriers when using either the Shannon or Simpsons diversity metrics (both p 0.2) (Fig. 1 B, C). We observed significantly lower Alpha diversity in S. aureus culture-positive samples than culture-negative samples when assessed with the Shannon diversity (p = 0.047) but not with the Simpson diversity (p = 0.09) (Fig. 1 D, E). We next investigated Beta diversity (similarity or dissimilarity between two samples) (Fig. 1 F-I) using Bray-curtis distance by colonisation status, which differed significantly (PERMANOVA analysis, F (2) = 36.67, p < 0.001). Distinct separation of samples by colonisation status could be clearly observed by non-metric multidimensional scaling (NMDS) (Fig. 1 F) and PCoA (Fig. 1 J) ordination plots. However, samples from intermittent carriers did not form a distinct cluster, and instead overlapped within the persistent or the non-carrier clusters, but with more samples from intermittent carriers being clustered with the non-carriers as visualised by the overlap in data ellipses in Fig. 1 F, G. This suggests that the microbiomes of intermittent (or rather occasionally S. aureus culture-positive) carriers are not distinct but typically more similar to non-carriers, with smaller numbers that have similar microbiomes to persistent carriers. Likewise, we observed similar distinct clusters between S. aureus culture-positive and culture-negative samples on the ordination plots (Fig. 1 H-I), as would be expected when S. aureus culture positivity is associated with persistent colonisation. Again, the two groups defined by S. aureus culture result differed significantly by PERMANOVA analysis ( F (2) = 59.01, p < 0.001) (Fig. 1 H, I). We only observed an association of sex with variation in the Bray-Curtis distance, as females (115/543, 21.2%) are less commonly persistent carriers compared to males (160/511, 31.3%) (p = < 0.001), but with a low F statistic and R 2 values ( F (2) = 2.83, p = 0.006, R 2 = 0.39%). There was no association with sex, smoking, pet ownership, healthcare worker, chronic skin condition, and diabetes. Compositional differences by colonisation status and defining community state types To visualise the causes of differences observed in Alpha and Beta diversity, we analysed species composition by S. aureus colonisation status (Fig. 1 J-K). The lower Alpha diversity of the persistent carriers was associated with the dominance of S. aureus in the species composition of this groups, compared to the ‘intermittent’ and non-carriers. In contrast, the nasal microbiome of non-carriers is largely dominated by multiple Corynebacterium species and D. pigrum . We next examined species composition at the level of each participant’s sample, separated by colonisation state (Fig. 1 L). This showed that amongst the 275 persistently colonised participants, S. aureus was the dominant organism (> 50% of reads) for 136/275 (49.5%), and in a subset of 96/275 (34.9%) participants, S. aureus represented > 75% of reads. In comparison, > 50% of reads from S. aureus was only seen in the 22/532 (4.1%), of S. aureus culture-negative (non-carriers) and 26/169 (15.4%) occasionally S. aureus culture-positive individuals (intermittent carriers). Instead, the non-carriers and subset of intermittent carriers were clearly dominated by three different Corynebacterium species ( C. pseudodipthericum, C. jeikeium, and C. accolens ) at abundances not seen in S. aureus persistent carriers (Fig. 1 L). Classifying individual samples by S. aureus culture result revealed that S. aureus was the predominant species (> 50% of reads) in 164/382 (42.9%) of the S. aureus culture-positive samples, and only 32/672 (4.76%) of culture-negative samples (Fig. 1 M). Men have higher S. aureus culture positive rates 35 , 36 so we examined the possibility of bias in culture by sex; however, we did not find that culture-negative samples with a higher S. aureus abundance (> 50% of reads) were more prevalent amongst females (17/32, 53.1%) compared to males (15/32, 46.9%). Community state types Next, we generated a heatmap of taxa abundance (Fig. 2 A), organised by hierarchical clustering by Bray-Curtis distance to examine the relationships between microbial residents of the anterior nares. We used this to define community state types (CSTs), i.e. samples with similar abundances of species which cluster together. To determine the number of clusters in the data, we calculated a gap statistic with ordination values using Bray-Curtis distances ( Supplementary Fig. S6 ). A total of 7 clusters were defined; we identified CSTs from the heatmap plot (Fig. 2 a). From the heatmap, it is evident that individuals always S. aureus culture-positive (persistent carriers) cluster to form the majority of CST1 (72.4%, 155/214), whilst those always S. aureus culture-negative (non-carriers) are represented largely by the remaining CSTs (Fig. 2 B-C). Intermittent carriers are dispersed across the CSTs. Using a multinomial logistic regression model, we found men had a reduced relative risk for association with CST6 (O.R.= 0.53, 95% C.I.= 0.30–0.93, p = < 0.05) and CST7 (O.R.=0.67, 95% C.I.= 0.47–0.96, p = < 0.05) compared with CST1 (Fig. 2 E). No other significant associations with CSTs were observed. Adjusted odd-ratios are provided in Supplementary Table S5 . We then formally evaluated differences in species abundances by colonisation status using ANCOM-BC2, which minimises the false discovery rate, using the unadjusted read count table. This demonstrated a significant differential abundance when comparing always S. aureus culture-negative individuals (non-carriers) and always culture-positive individuals (persistent carriers) carriers, with a positive association of S. aureus seen with persistent carriage (as expected), and a negative association seen with multiple Corynebacterium species, D. pigrum , S. epidermidis , and M. catarrhalis (Fig. 2 F and Supplementary Table S4 ). No significant differences in species abundance other than S. aureus between non-carriers and intermittent carriers is observed (Fig. 2 F and Supplementary Table S4 ). Notably, persistent carriers had a greater log-fold change in S. aureus when compared with occasionally S. aureus positive individuals (intermittent carriers), in comparison to non-carriers, suggesting the relative abundance of S. aureus may be driving its longitudinal carriage. In a subgroup of 34 participants, from the rarefied dataset to 10,000 reads, two or three samples (n = 75) were available over consecutive weeks ( Supplementary Fig. S7-10 ). These included 13 persistent carriers, 7 intermittent carriers, and 14 non-carrier’. We examined the stability of the community in the anterior nares and correlated pairwise Alpha diversity of participants by colonisation status. Persistent (Spearman’s rho = 0.54, p = 0.028) and intermittent carriers (Spearman’s rho = 0.79, p = 0.028) were found to have greater stability compared to non-carriers (Spearman’s rho = 0.30, p = 0.268). Further examining the microbiome of ‘intermittent’ carriage Having observed that the majority of intermittent carriers clustered with S. aureus non-carriers group (e.g. overlapping data ellipses in Fig. 1 F, G), we hypothesised that intermittent carriers could be misclassified non- or persistent carriers. We examined differences in Alpha diversity between the one and two swab positive intermittent subgroup ( Supplementary Fig. S11 ), using only samples with greater than 10,000 reads. We found no significant difference in Alpha diversity when comparing samples with one S. aureus positive swab compared with two (p = 0.21). Beta diversity by Bray-Curtis index between samples with one or two positive S. aureus swabs did differ significantly by PERMANOVA analysis ( F (2) = 3.19, p = 0.003), suggesting that these groups have differing microbial compositions (Fig. 3 A-D). We next explored the abundance of species across the samples depending on the number of swabs which were positive for S. aureus . There is a clear continuous trend in the variation in abundance from zero to three positive swabs (Fig. 3 F). We then subset the participants representing intermittent carriers (n = 169) from the dataset to examine if these were two distinct populations (rather than one) based on the number of S. aureus positive swabs (one swab, n = 103 and two swabs, n = 66). From examination of the species composition of individual samples, a different microbial community structure is apparent for intermittent carriers who are positive for two swabs compared to those with one swab (Fig. 3 F). We formally analysed the differences in community structure using a heatmap of abundances from the samples of intermittent carriers, which displays a similar structure of clustering to that observed when comparing persistent carriers (Fig. 3 G). Again, we calculated a gap statistic, giving an optimal number of CSTs of 7 (same as full dataset), and the hierarchical clustered dendrogram was split accordingly (Fig. 3 G). On this heatmap, it is clear that the CSTs that are dominated by Corynebacterium species, D. pigrum and S. epidermidis are associated with samples where participants had one positive S. aureus culture. 18/66 (27.3%) intermittent carriers who had two positive S. aureus cultures were associated with a CST dominated by S. aureus , compared to 7/103 (6.8%) with one positive swab. These findings reflect similar observations for persistent carriers and non-carriers, respectively (Fig. 1 F, G), and provide further evidence that ‘intermittent carriers’ are just misclassified persistent carriers or non-carriers. Predicting colonisation status from the nasal microbiome We next used a random forest model to establish whether microbiome data could be used to predict the culture-based categorisation of nasal S. aureus colonisation status. Additionally, this served as a sensitivity analysis for the previous differential abundance analysis (Fig. 2 F), which allows for the identification of significant microbial determinants for S. aureus colonisation status. We split the data into training and test data at a ratio of 80:20, and determined the best number of candidates to be sampled at each tree (mtry) to be 6. The estimated test classification accuracy of the trained model was 73.2% (1-estimated out of box error) with the lowest class error for non-carriers (6.85%) and highest for intermittent carriers (100%). We determined the accuracy, sensitivity and specificity of the model with the test data. The overall accuracy of the model was 75.2% (C.I.= 67.4%-81.9%, p = < 0.001) significantly exceeding the no information rate (Fig. 4 A). Overall, the model performed best in predicting persistent colonisation with 83.0% and 88.2% sensitivity and specificity, respectively, suggesting the greatest utility for identification of individuals at higher risk of persistent S. aureus colonisation (Fig. 4 B). For non-carriers, the sensitivity was higher at 94.8%, but specificity lower at 66.6%. For intermittent carriers the sensitivity was 0.0% suggesting the model was completely unable to predict the intermittent colonisation from the microbiome data; of the 25 intermittent carriers in the test dataset, none were classified as intermittent carriers, 16/25 (64%) were misclassified as non-carriers and 9/25 (36%) as persistent carriers, adding further evidence that intermittent carriers are not distinct group, and a greater proportion are similar to non-carriers compared to persistent carriers. We determined variable importance (i.e. how much each variable contributes to the prediction) by evaluating the mean decrease in accuracy (a measure of decrease in the model accuracy computed by permuting out-of-box error data) and the mean decrease in gini index (a measure of variance and resulting misclassification across the random forest nodes) after removal of each feature, i.e. taxon 37 . The top three features of importance by assessing the mean decrease in accuracy were S. aureus , Corynebacterium sp. , and S. epidermidis (Fig. 4 C). The top three features of importance by assessing the mean decrease in gini index were S. aureus , Corynebacterium sp. , and D. pigrum , with S. aureus clearly contributing the most to the model (Fig. 4 D). Staphylococcus aureus phylogenetic associations with carriage We next investigated if certain S. aureus lineages have a propensity for persistent nasal carriage or are more capable of dominance of the community compared to other competing resident bacteria. We used S. aureus isolate whole genome sequences with matched microbiome data (n = 172) and compared the S. aureus phylogenetic tree, and major multi-locus sequence types (MLST), to the colonisation state and the sample microbiome (Fig. 5 A). Following the bifurcation at the root of the phylogeny, as seen in large collections of diverse S. aureus 38 , there is a greater number of samples showing higher S. aureus abundance amongst isolates in cluster B (dominated by ST30, ST34, ST398, and ST45) with a lower number of samples showing higher abundance of species identified earlier as showing a negative association with S. aureus (Fig. 2 F). We examined differences in rarefied S. aureus abundance (n = 125), which demonstrated a significantly higher abundance in samples in cluster B compared to cluster A (Mann-Whitney, p = 0.041) (Fig. 5 B). For the S. aureus dominant CST1 (Fig. 2 ), 38/74 (51.4%) of cluster A compared to 33/51 (64.7%) cluster B samples were found in this CST. Next, we assessed differences in Beta diversity between cluster A and B (Fig. 5 C, D), and found a small but statistically significant (PERMANOVA analysis ( F (2) = 2.33, p = 0.04) divergence of these groups. Discussion Despite the importance of S. aureus colonisation as a risk factor for S. aureus infection, there is still only a limited understanding of what determines nasal S. aureus colonisation. In this work, we combine for the first time, large-scale microbiome sequencing with longitudinal culture data that, since the 1940’s 32 , has been used to define S. aureus colonisation. We have generated multiple new insights into the nasal microbial community structure of the anterior nares, substantially extending previous smaller-scale studies 3 , 17 , 31 , 39 , 40 . Like the previous study of older twins from Denmark 41 , we identify seven community state types (CSTs), but with a different species composition which suggests that either the previous smaller study was unrepresentative or there is variation in the nasal microbiota even between two northern European countries. Importantly, our analysis of the seven CSTs revealed that women are more likely to have either CST 6 ( C. accolens ) or CST 7 (diverse group), suggesting an influence of sex on the wider nasal microbiome composition, as with S. aureus persistent colonisation 12 and load 41 . We demonstrate that there is a clear distinction in the microbial community structure that underlies ‘persistent’ S. aureus carriage compared to non- S. aureus carriage in a large sample of individuals (Fig. 5 E). We found that persistent carriage of S. aureus is negatively associated with three Corynebacterium species (including C. jeikeiu m, C. accolens and an unnamed Corynebacterium sp ), D. pigrum , S. epidermidis , and M. catarrhalis . Notably, C. jeikeium , C. accolens, M. catarrhalis , and the unnamed Corynebacterium sp. have not been previously identified as negatively associated with S. aureus abundance in microbiome data 31 , 41 , 42 . We failed to replicate the negative association with S. aureus abundance with Simonsiella sp. or Cutibacterium (formerly Propionibacterium) granulosum as previously reported 41 . Neither species were found in our pre or post-QC data, suggesting that these are either uncommon species in England, perhaps only present in certain environmental conditions, or contaminants 29 . In contrast, our findings did replicate the negative association previously reported for Dolosigranulum spp. 41 and D. pigrum 16 . Interactions between D. pigrum and Corynebacterium spp. and particularly C. accolens have recently been explored in vitro ; D pigrum was shown to inhibit a single strain of S. aureus directly, whilst C. pseudodipthericum , C. accolens , C. propinquum conditioned media enhanced the growth of D. pigrum , and that this growth enhancement was not reciprocal 16 . C. accolens was found to both enhance the growth of D. pigrum through an unknown mechanism(s), and inhibit growth by processing host tri-acylglycerols into fatty acids with antibacterial properties 16 . Our data further highlights the extent of the mutualism between taxa that are negatively associated with S. aureus carriage such as D. pigrum and multiple Corynebacterium spp. (see co-occurrence in CST2, 3 and 4 in Fig. 2 a), suggesting that this is an area which needs further systematic investigation. We found both C. jeikeium and C. accolens are negatively associated with S. aureus . While C. jeikeium has not been previously reported to inhibit S. aureus , C. accolens has been demonstrated to inhibit S. aureus experimentally 43 , 44 , and as noted above, C. accolens can support or inhibit growth of D. pigrum - which can inhibit S. aureus directly 16 . This suggests C. accolens can inhibit S. aureus through direct or indirect means. This is supported by our data whereby most individuals with C. accolens are found in CST6, which is dominated by C. accolens and has a low abundance of S. aureus (Fig. 2 a). Notably, women have an increased relative risk for CST6; this provides a mechanistic explanation for why women are less likely to be persistent carriers of S. aureus , although this is unlikely to explain the entire variation in colonisation rates between men and women 12 . As discussed previously, C. accolens has also been reported to be positively associated with S. aureus in a smaller microbiome study and was shown to promote the growth of S. aureus in vitro 3 . In our data, C. accolens is found at a lower abundance in multiple CSTs, including co-occurrence with S. aureus in a small sub-cluster within the S. aureus dominated CST1 (see CST1 in Fig. 2 a). This suggests that there is a strain level or lineage variation in C. accolens and/or S. aureus that contributes to this discordant relationship – which again highlights the need for large scale studies. Our data identified a negative association between S. epidermidis and S. aureus , corroborated by the differential abundance analysis and the random forest model. Although negative 31 and positive 41 associations have been reported previously, similar to C. accolens , a previous study has shown that a certain proportion of S. epidermidis strains secrete a serine protease, Esp, which inhibits S. aureus growth 19 . Furthermore, the effect observed could also be explained by environmental factors. S. aureus requires a higher relative humidity (87%) than S. epidermidis (81%), and S. epidermidis may be more commonly found in "drier" noses 45 . Our finding that M. catarrhalis is negatively associated with S. aureus in adults is new, though abundance has been reported to be inversely correlated with S. aureus in children 18 . As would be expected, persistent carriage was positively associated with S. aureus . Strikingly, we found that in ~ 50% of persistently colonised individuals S. aureus is the single most abundant organism in the nasal microbiome, representing > 75% of reads for ~ 35% persistent carriers, indicating that when present it dominates the niche. This domination is reflected in reduced Alpha diversity among persistent carriers compared to non-carriers and is further supported by the greater stability of Alpha diversity amongst the persistent S. aureus carriers compared to non-carriers. This suggests that S. aureus may act as a keystone species which principally determines its own continued carriage and can suppress other members of the nasal microbiome. While most nasal microbiome research has concentrated on how other members of the nasal microbiome prevent S. aureus colonisation, our data indicates that future investigations should now focus how S. aureus excludes other species from the nasal microbiome and is resistant to antagonistic compounds or growth conditions generated by competing species such as Corynebacterium spp. and D. pigrum . The subset of very high S. aureus load carriers we identified might also be of clinical importance: The high bacterial load of S. aureus and a lack of antagonistic species amongst persistent carriers specifically may explain why these individuals are more likely to become infected by S. aureus ; inoculation of a wound with a high load of S. aureus in the absence antagonistic species (that might inhibit S. aureus ) could make infection is more likely. Future studies which make the use of nasal microbiome-based stratified participants are required to understand the consequences of high-load S. aureus dominant nasal colonisation on the rest of the human microbiome (e.g. skin and gut 26 ) and subsequent risk of infection. Importantly, our study demonstrates that S. aureus carriage can be predicted from microbiome data with a moderate degree of accuracy. Notably, the model is more sensitive in predicting S. aureus non-carriage. This is particularly important given a single swab of the anterior nares is limited in its diagnostic accuracy 33 . Further, the high negative predictive value of the model for persistent carriage we present here, using the microbiome data from a single swab, may improve the identification of true negatives in S. aureus screening and identify those who are unlikely to be persistently colonised, facilitating a selective approach to patient decolonisation. Further large-scale studies using higher resolution metagenomics and clinical data will likely significantly improve this, moving towards being able to define risk of S. aureus infection based on a single swab. We have determined that ‘intermittent’ carriers do not have a distinct microbial community. This was first proposed by Van Belkum et al . 10 , 46 who theorised that: (1) non-carriage is either incidental and most people are actually intermittent carriers or (2) intermittent carriers are non-carriers who have picked up S. aureus from the environment. Our data provides evidence that the latter of the two is partly correct and provides new insights into this. Intermittent carriers, as defined by one or two culture positive swabs, belong to one of two populations: a population with a S. aureus dominated microbial community structure (similar to persistent carriers) with the absence of species identified as negatively associated with S. aureus or a population with one of several microbial community structures which are not dominated by S. aureus , are more diverse, and often dominated by other species (similar to non-carriers) (Fig. 5 e). Therefore, the most parsimonious explanation is that ‘intermittent’ carriers, given their low S. aureus abundance and predominance amongst non-carriage CSTs, are effectively S. aureus ‘non-dominant carriers’ who are only transiently colonised with S. aureus , for example due to environmental/household exposure (i.e. hypothesis 2 proposed by Van Belkum et al .). While two swab positive ‘intermittent’ carriers are ‘ S. aureus dominant carriers’ (akin to persistent carriers), reflected by their higher S. aureus abundance and greater representation amongst the S. aureus dominant CST, and therefore likely to be persistently colonised individuals that were negative by culture in one swab. We identified a relationship between certain S. aureus lineages and S. aureus abundance and the associated microbiome. This suggests that S. aureus abundance and carriage is to some degree lineage specific. Previous studies of human experimental colonisation with S. aureus identified that after decolonisation and artificial inoculation, persistent carriers had higher loads of S. aureus than intermittent or non-carriers and were more likely to select their own strain 10 . Our data suggest that this might be due to lineage-specific effects of persistent carriers’ strains being better adapted to persistent colonisation, enabling them to reach higher abundances. The same study also showed that persistent carriers had higher serum IgG and IgA levels to certain S. aureus antigens (SasG, TSST-1, SEA, ClfA, CHIPS). Given the known variation of mobile genetic element (MGE) content in S. aureus lineages (TSST-1, SEA, CHIPS are all MGE acquired), a propensity for lineages with particular MGE content to be found at higher abundance in persistent carriers may explain this variation in antibody levels. Overall, these lineage relationships are likely to represent adaptations of S. aureus 47 that impact host adaption to colonisation, transmission success and intra- and interspecies competition. This requires further investigation to further elucidate the mechanisms for this lineage specific adaption to colonisation. This study has potential limitations, including those inherent to 16S rRNA gene studies; though we have previously demonstrated the accuracy of our sequencing and analysis pipeline 48 , further validation of taxa (for example: the unnamed Corynebacterium sp. negatively associated with S. aureus ) identified at a species level with selective culture and/or shotgun metagenomic sequencing will be useful. Additionally, although the study participants are healthy and sampled at home, the blood donor cohorts used will still not be entirely representative of the population of England, which is itself clearly not representative of all global populations. Further studies in different populations using standardised methods are now required to explore this variation. In summary, we present the most comprehensive assessment of the microbial composition of the anterior nares to date. Our data provides multiple new insights and identifies key microbial interactions and variation that underpin the composition of the human nasal microbiome, and in particular colonisation by S. aureus. Methods Ethics approval and consent to participate The CARRIAGE study protocol was approved by the National Research Ethics Service Committee North-West - Lancaster Research Ethics Committee, 27/06/2016, REC reference: 16/NW/0507, IRAS project ID: 202688. All participants provided informed consent. The study is registered at ISRCTN: ISRCTN10474633 . Participants and samples In this observational cohort study, nasal samples taken from the anterior nares were obtained from healthy human participants from the community participating in the CARRIAGE study 49 between 13 th October 2016 and 17 th May 2017. Briefly, S. aureus colonisation status was assessed by culture of three self-administered nasal swabs delivered to participants and taken at weekly intervals, and subsequently posted back to the laboratory. S. aureus colonisation status was defined as: (i) persistent colonisation, based on three S. aureus culture positive weekly nasal swabs, (ii) intermittent colonisation, defined as one or two swabs positive, and (iii) non-carrier status, defined as no swabs positive, based on previous studies 9,10,33,34 ( Fig. 1A ). Lifestyle information was collected by questionnaires or from pre-existing data held as part of baseline questionnaires in previous studies involving the same participants. The Amies transport liquid that the swabs (the same swabs that were used for culture) were transported to the laboratory in were processed without culture for 16S rRNA gene sequencing to identify the microbial community composition ( Supplementary Fig. S1 and 2 ). S. aureus culture After 10 seconds of vortexing, nasal swabs in Amies transport media (Medical Wire) were transferred to a tube containing 2 ml Tryptic Soy broth supplemented with 6.5 % NaCl (Medical Wire) and incubated overnight at 37 °C, in air. The remaining Amies solution was transferred to an Eppendorf with 500 µl of glycerol, pipette mixed and stored at -70ºC. 10 µl of the overnight enrichment broth was streaked onto chromogenic Staph Brilliance 24 agar plates (Oxoid) and incubated overnight at 37 °C. If no blue colonies were identified after 24 hours of incubation, the plate was returned to the incubator overnight and rechecked. Blue colonies are with the phenotypes of putative S. aureus were sub-cultured onto Columbia Blood agar (5% horse blood) and incubated overnight at 37 °C. Colonies from these plates were inspected visually for phenotype indicators, and tested for coagulase and protein A via latex agglutination test (Pro-Lab Diagnostic). Where there were queries or discrepancies, species level identity was confirmed using Matrix assisted laser desorption and ionisation – Time of Flight (MALDI-ToF). All isolates were stored (Pro-Lab Diagnostics) at -80 °C. DNA processing and 16S rRNA gene polymerase chain reaction Prior to extraction, residual sample transport medium from nasal samples was stored at -70ºC in ~33% v/v glycerol. Total DNA was extracted from nasal sample transport medium after an additional mechanical lysis step (Lysing matrix E, MP Biomedicals) either via the MPBio MPure-12 instrument, (MPure Bacterial DNA Kit, MP Biomedicals) or manually using the FastSpin Kit for Soil (MPBiomedicals), including the heated elution step. DNA was then stored at -70ºC until use. V1V2 specific primers with attached sequencing adaptors and indexes ( Table S1 ) were used for PCR to amplify the bacterial 16S ribosomal gene regions. PCR amplification mastermixes were prepared manually using a Q5 High-Fidelity Polymerase Kit (M0491, New England Biolabs). PCR amplifications were setup in triplicate (25ul each), products were pooled into a single volume per sample, and all samples were subsequently purified using an AMPure XP (Beckman Coulter) workflow at a ratio of 0.8X. Libraries were quantified using the Qubit HS DNA Kit (ThermoFisher). Equimolar pools were then created. Negative controls included a sample extraction control, a PCR water control, and an aliquot of the glycerol used for storage, whilst a positive control was represented by purified water spiked with S. aureus DNA. DNA sequencing Per experiment, an equimolar pool of PCR libraries was sequenced at the Wellcome Sanger Institute in-house sequencing facility, using the Illumina MiSeq (300bp paired-end reads, v3 Reagent Kit). Accession numbers for the sequencing data is in Supplementary Table S6 . 16S rRNA gene sequence quality control and taxonomy assignment We used a modified mothur MiSeq standard operating procedure (SOP) to process paired fastq files (MOTHUR wiki at http://www.mothur.org/wiki/MiSeq_SOP) 50 . The four poly(NNNN)s present in the adapter/primer sequences of contigs assembled with the ‘make.contigs’ command in mothur were trimmed with the PRINSEQ program, before the modified MiSeq SOP was resumed. The Silva bacterial database ‘silva.nr_v132.align’ was used to align quality-screened sequences and chimeras removed using Uchime 51 . Sequences were then classified using the same Silva reference database and the Silva taxonomy database ‘silva.nr_v132.tax’, with the removal of chloroplast, mitochondria, unknown, and eukaryota sequences. We clustered high-quality unique sequences with Oligotyping v2.1 52 (-M option to 1000), which were assigned to NODES, and referred to as operational taxonomic units (OTU) from here, with the ‘Minimum Entropy Decomposition’ (MED) option. We created a customised silva SSU Ref database (NR99, release 132), where we removed the majority of environmental and uncultured taxa, and carried out taxonomic assignment with ARB (v6.0.6-3) 53 . In some instances, where a mismatch was observed within the taxonomic groups, we assigned taxa to the OTU sequence with BLAST 54 (see Supplementary Table S2 ). We then combined the output in R (v4.4.1) into a phyloseq 55 object for onward analysis. Contaminant removal and accounting for variability in sequencing depth We identified contaminants and removed these by identifying batch effects and accounting for negative controls 28-30 . Batch effects were assessed by calculating the spearman’s correlation co-efficient of species against each location of extraction, and location of PCR reaction. We then examined correlation of species with sample DNA concentrations. We used well characterised ‘kitome’ and environment contaminants to identify additional associated contaminants by calculating ‘species-species’ correlation coefficients. We used Decontam v1.16.0 56 to account for laboratory negative controls, run with the ‘isnotcontam’ function and with each sequencing run provided as a batch (further details below and in Supplementary Fig. S3 and S4 and Supplementary Table S3 ) We determined a suitable rarefication depth of 10,000 reads using rarefication curves and examining the read depth at which the majority of sample taxa numbers plateaued (see rarefaction section below and Supplementary Fig. S5 ). We removed species with an abundance of less than 0.1% across samples, below which we expected the removal of most contaminants and account for the variability in rare species composition between runs 48 . For diversity analyses, the rarified dataset was used. For abundance analyses, to mitigate data loss, we combined samples with greater than 500 high quality reads with samples that had greater than 10000 reads and rarefied. (see rarefaction section, Supplementary Fig. S3, S4, S5 and Supplementary Table S3 ). Identification of contaminants Removal of taxa below the 0.1% threshold resulted in 115/2,322 OTUs remaining. The samples were processed over two time periods. Over the first time period (n=1,099), there were five locations for DNA extractions and five locations for PCR amplification. we used spearman’s correlation coefficient to identify batch effects, specifically species with abundance that was associated with the extraction and PCR locations ( Supplementary Table S3, Supplementary Fig., S3 and S4 ). For the second time period (n=767), extractions and PCR amplifications took place in one location and therefore batch effects by location was not examined. We used spearman’s correlation coefficient to identify taxa that correlated with PCR qubit values (post-PCR amplification DNA concentration); previously, lower sample DNA concentrations have been associated with contaminants 30,57 ( Table S3, Supplementary Fig., S3 and S4 ). We used hierarchical clustering to identify species that clustered with one another, which allowed for the identification of taxa that were correlated with well-characterised and suspected contaminants 30 ( Supplementary Table S3, Supplementary Fig., S3 and S4 ). As a final check, we used the R package Decontam (v1.16.0) 56 to account for negative controls, with each sequencing run considered as a batch ( Supplementary Table S3, Fig. S3 and S4 ). Determining a rarefication threshold We subset CARRIAGE samples and generated rarefication curves for samples with greater than 1,000, 5,000, 10,000, 15,000, 20,000, 100,000 high-quality reads respectively ( Supplementary Fig. S5 ). In order to determine a rarefication threshold, we identified the slope of each rarefication curve at the respective high-quality read threshold using the rareslope() function in phyloseq ( Supplementary Fig. S5 ); given the large dataset, visualising the point at which the curves plateaued was not possible. It was apparent that at greater high-quality read thresholds, a larger proportion of the samples reached a (near) plateau. We aimed to reach a balance between losing a large number of samples and retaining a dataset where the rarefication curves for the vast majority of samples had plateaued; this was met at 10,000 reads ( Supplementary Fig. S5 ). From here, we either use the dataset rarefied to an even-depth to the minimum read count above this threshold (10,004) or this dataset combined with samples with greater than 500 reads but less than 10,000 reads, to minimise data loss and consistent with previous analyses 58 . Diversity analysis We conducted microbial diversity and compositional analysis in R using diversity indices calculated with the phyloseq (v1.40) 55 and vegan (v2.6-4) 59 packages. Alpha-diversity indices (Shannon’s and Simpson’s) were calculated on rarefied read counts. Sample microbial composition is consistently represented with relative abundances. We used Principal Coordinate Analysis (PCoA) and Non-Metric Dimensional Scaling (NMDS) with the bray–curtis distance matrices to visualise differences in sample diversity by condition (e.g. S. aureus colonisation status). Data visualisation and statistical analysis We manipulated data in Excel 2016 and R (v4.4.1). We generated figures using ggplot2 (v3.4.0), phyloseq (v1.40) 55 , and microViz (v0.11.0). We evaluated differences in Alpha indices with Mann-Whitney-U and Kruskall-Wallis tests where appropriate. We used PERMANOVA to estimate differences between Bray-Curtis distances observed by study groups with the vegan package (v2.6-4) 59 . To determine the number of clusters in the data, we calculated a gap statistic with ordination values using Bray-Curtis distances, using the R package ‘cluster’ function clusGap() ( Fig. S6 ). We investigated the association of plausible lifestyle and comorbidities risk factors with Community State Types (CST) using a multinomial logistic regression model analysis (CST ~ sex + smoking status + pet ownership + healthcare contact + chronic skin condition + asthma + allergies). We used ANCOM-BC (v1.6.4) 60 to evaluate differential abundance of microbial species in the study groups; we used the ancombc2 function with default settings, but specified taxa with a prevalence of less than 0.1% to be removed and a library cut-off of 500 reads, and provided a non-rarefied count table as a centred log ratio transformation is conducted 60 . This study complies with the STORMS guidelines 61 for reporting. Random forest model We trained two separate models, one utilising all samples above 500 reads where samples with greater than 10,004 had been rarefied (n=1,055), and another including the rarefied dataset alone (n=795). The rarefied dataset performed better compared to the combined dataset (see Supplementary Results for further details). We used the R package randomForest (v4.7-1.1) 37 to fit a random forest classifier for carriage status (relative_microbial_abundance_data ~ carriage_status). The model was trained using a randomly subsampled dataset of the microbial features (in relative abundance format) representing 80% of the data (ntrees=1000), and tested on the remaining 20% to evaluate model robustness. We chose the number of predictors sampled for splitting at each node (mtry) with the tuneRF() function. We obtained sensitivity and specificity values of the model with the R package caret (v6.0-90) whilst receiver operating characteristic curve (ROC) curves and AUC were obtained with the R package pROC (v1.18.4). P-values less than 0.05 were considered statistically significant. Whole Genome Sequencing of Staphylococcus aureus isolates S. aureus isolates were sequenced at the Wellcome Sanger Institute with 96 sample libraries sequenced on a 300bp PE MiSeq lane (with a 1% PhiX spike). European Nucleotide Accession number for isolates is presented in Supplementary Table S7 . Phylogenetic analysis From the raw whole genome sequencing data, we generated quality control metrics, and trimmed reads, with the nextflow pipelines, bacQC (github.com/avantonder/bacQC). Species classification for each sample was performed using Kraken and Bracken 62 . We discarded samples with less than 90 % reads matching to S. aureus and those with <30x coverage from onward analyses. Using the nextflow pipeline, assembleBAC (github.com/avantonder/assembleBAC), we produced annotated assemblies with trimmed fastqs. The pipeline uses shovill (v1.1.0) for assembly. We annotated assemblies with prokka (v.1.14.5) 63 using a genus-specific database from RefSeq for annotation. Assemblies with an N50 value <10000, length of less than 2.6Mbp or greater than 3.0Mbp, or with a spuriously high number of contigs summarised by MultiQC 64 and the QC metrics generated by Panaroo(v1.3.4) 65 were removed from onward analyses. Samples with greater than 300 contigs were found to be outliers. We assigned sequence types (STs) with mlst (v2.19.0) (github.com/tseemann/mlst), and where these were not assigned, assemblies we queried the sequences on the PubMLST web server 66 . We produced core-genome alignments with Panaroo (v 1.3.4) 65 with a core-genome threshold set to 98 %. We extracted variant sites from the core-genome alignment with snp-sites (v2.5.1) 67 and coupled with associated values for invariant sites to build a maximum likelihood (ML) phylogenetic tree. We used IQ-TREE (v2.1.2) 68 to estimate ML phylogenetic trees with the optimal phylogenetic trees determined by ModelFinder 69 and branch support statistics generated using the ultrafast bootstrap method 70 . Declarations Data availability All sequencing data is publicly available in the European Nucleotide Archive, with details outlined in Supplementary Table S6 and S7. Code availability The authors declare that all data cleaning and analysis associated with this article were performed using previously published methods, the applications of which are appropriately cited in the corresponding sections in the Methods. No custom code was developed for the aforementioned purposes. Additional code underlying the figures featured are available from the corresponding authors upon request. Acknowledgements We wholehearted thank the CARRIAGE study participants for taking part in the CARRIAGE study and providing the samples that were critical to being able to conduct this research. Funding This work was supported by Wellcome Collaborative Award in Science (Grant no. 211864/Z/18/Z) to SJP, JP, JD, JAG, EMH. Isaac Newton trust Grant 17.07(1) to EMH, UKRI Innovation Fellowship: MR/S00291X/1 to EMH, Wellcome Grant reference: 220540/Z/20/A, 'Wellcome Sanger Institute Quinquennial Review 2021-2026' – core funding of Wellcome Sanger Institute, Wellcome Clinical PhD Fellowship: 222903/Z/21/Z to DAg. This research was supported by the NIHR Cambridge Biomedical Research Centre (NIHR203312*). The epidemiological coordinating centre of the CARRIAGE study was additionally supported by awards from the NIHR Blood and Transplant Research (5Unit (BTRU) in Donor Health and Behaviour (NIHR203337), NIHR Cambridge BRC (NIHR203312) (*), and by Health Data Research UK (HDRUK2023.0028), which is funded by the UK Medical Research Council, Engineering and Physical Sciences Research Council, Economic and Social Research Council, Department of Health and Social Care (England), Chief Scientist Office of the Scottish Government Health and Social Care Directorates, Health and Social Care Research and Development Division (Welsh Government), Public Health Agency (Northern Ireland), British Heart Foundation and Wellcome. J.D. holds a British Heart Foundation Personal Chair (CH/12/2/29428). *The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care'. Authors' contributions Conceptualisation: SJ, EMH, JP, JD, JAG, and DAg. Methodology: DAg, MCdG, KLB, JW, JP, SJP, EMH Data curation: DAg, KLB, BB, SB, RH, CP, MRW, CM, SG, CRdS, LS, JB, SD, EJ, MJ, DAn, SI, AM. Investigation: KLB, BB, KE, PN, SG, CRdS, LS, LL, CR, XB, JB, DAg. Software: DAg, MCdG, AvT. Resources: JD, AB, ED, MH, SJP, EMH Formal analysis: DAg. Validation: DAg, MCdG. Visualization: DAg, DYKN. Writing—original draft preparation: DAg, EMH Writing—review and editing: All authors. Project administration: KLB, BB, DAg, SB, RH, CP, MRW, CM, CC, SD, EJ, MJ, DAn, SI, AM, SJP, EMH. Supervision: EMH, SJP, JP, JD, MCdG, JW. Funding acquisition: EMH, SJP, JP, JD, JAG, DAg. Ethics declarations Competing interests A.B. reports institutional grants from AstraZeneca, Bayer, Biogen, BioMarin, Bioverativ, Novartis, Regeneron and Sanofi. J.D. serves on scientific advisory boards for AstraZeneca, Novartis, and UK Biobank, and has received multiple grants from academic, charitable and industry sources outside of the submitted work. References Moss B, Squire JR (1948) Nose and skin carriage of Staphylococcus aureus in patients receiving penicillin. Lancet 1:320–325. https://doi.org:10.1016/s0140-6736(48)92088-1 Reagan DR et al (1991) Elimination of coincident Staphylococcus aureus nasal and hand carriage with intranasal application of mupirocin calcium ointment. Ann Intern Med 114:101–106. https://doi.org:10.7326/0003-4819-114-2-101 Yan M et al (2013) Nasal microenvironments and interspecific interactions influence nasal microbiota complexity and S. aureus carriage. Cell Host Microbe 14:631–640. https://doi.org:10.1016/j.chom.2013.11.005 Kluytmans J, van Belkum A, Verbrugh H (1997) Nasal carriage of Staphylococcus aureus: epidemiology, underlying mechanisms, and associated risks. Clin Microbiol Rev 10:505–520. https://doi.org:10.1128/CMR.10.3.505 Wertheim HF et al (2004) Risk and outcome of nosocomial Staphylococcus aureus bacteraemia in nasal carriers versus non-carriers. Lancet 364:703–705. https://doi.org:10.1016/s0140-6736(04)16897-9 von Eiff C, Becker K, Machka K, Stammer H, Peters G (2001) Nasal carriage as a source of Staphylococcus aureus bacteremia. Study Group. N Engl J Med 344:11–16. https://doi.org:10.1056/nejm200101043440102 Bode LG et al (2010) Preventing surgical-site infections in nasal carriers of Staphylococcus aureus. N Engl J Med 362:9–17. https://doi.org:10.1056/NEJMoa0808939 Wertheim HF et al (2005) The role of nasal carriage in Staphylococcus aureus infections. Lancet Infect Dis 5:751–762. https://doi.org:10.1016/s1473-3099(05)70295-4 Nouwen JL et al (2004) Predicting the Staphylococcus aureus nasal carrier state: derivation and validation of a culture rule. Clin Infect diseases: official publication Infect Dis Soc Am 39:806–811. https://doi.org:10.1086/423376 van Belkum A et al (2009) Reclassification of Staphylococcus aureus nasal carriage types. J Infect Dis 199:1820–1826. https://doi.org:10.1086/599119 Cole AL et al (2018) Cessation from Smoking Improves Innate Host Defense and Clearance of Experimentally Inoculated Nasal Staphylococcus aureus. Infect Immun 86. https://doi.org:10.1128/IAI.00912-17 Sollid JU, Furberg AS, Hanssen AM, Johannessen M (2014) Staphylococcus aureus: determinants of human carriage. Infect Genet evolution: J Mol Epidemiol evolutionary Genet Infect Dis 21:531–541. https://doi.org:10.1016/j.meegid.2013.03.020 Scheuch M et al (2019) Staphylococcus aureus colonization in hemodialysis patients: a prospective 25 months observational study. BMC Nephrol 20:153. https://doi.org:10.1186/s12882-019-1332-z Nguyen MH et al (1999) Nasal carriage of and infection with Staphylococcus aureus in HIV-infected patients. Ann Intern Med 130:221–225. https://doi.org:10.7326/0003-4819-130-3-199902020-00026 Mulcahy ME, McLoughlin RM (2016) Host-Bacterial Crosstalk Determines Staphylococcus aureus Nasal Colonization. Trends Microbiol 24:872–886. https://doi.org:10.1016/j.tim.2016.06.012 Brugger SD et al (2020) Dolosigranulum pigrum Cooperation and Competition in Human Nasal Microbiota. mSphere 5 https://doi.org:10.1128/mSphere.00852-20 Liu CM et al (2015) Staphylococcus aureus and the ecology of the nasal microbiome. Sci Adv 1:e1400216. https://doi.org:10.1126/sciadv.1400216 Accorsi EK et al (2020) Determinants of Staphylococcus aureus carriage in the developing infant nasal microbiome. Genome biology 21, 301 http://europepmc.org/abstract/MED/33308267 https://genomebiology.biomedcentral.com/track/pdf/10.1186/s13059-020-02209-7 https://doi.org/10.1186/s13059-020-02209-7 https://europepmc.org/articles/PMC7731505 https://europepmc.org/articles/PMC7731505?pdf=render Iwase T et al (2010) Staphylococcus epidermidis Esp inhibits Staphylococcus aureus biofilm formation and nasal colonization. Nature 465:346–349. https://doi.org:10.1038/nature09074 Zipperer A et al (2016) Human commensals producing a novel antibiotic impair pathogen colonization. Nature 535:511–516. https://doi.org:10.1038/nature18634 Bomar L, Brugger SD, Lemon KP (2018) Bacterial microbiota of the nasal passages across the span of human life. Curr Opin Microbiol 41:8–14. https://doi.org:10.1016/j.mib.2017.10.023 de Piters S, Binkowska WAA, J., Bogaert D (2020) Early Life Microbiota and Respiratory Tract Infections. Cell Host Microbe 28:223–232. https://doi.org:10.1016/j.chom.2020.07.004 Teo SM et al (2015) The infant nasopharyngeal microbiome impacts severity of lower respiratory infection and risk of asthma development. Cell Host Microbe 17. https://doi.org:10.1016/j.chom.2015.03.008 Biesbroek G et al (2014) Early respiratory microbiota composition determines bacterial succession patterns and respiratory health in children. Am J Respir Crit Care Med 190:1283–1292. https://doi.org:10.1164/rccm.201407-1240OC Uehara Y et al (2000) Bacterial interference among nasal inhabitants: eradication of Staphylococcus aureus from nasal cavities by artificial implantation of Corynebacterium sp. J Hosp Infect 44:127–133. https://doi.org:10.1053/jhin.1999.0680 Piewngam P et al (2018) Pathogen elimination by probiotic Bacillus via signalling interference. Nature 562:532–537. https://doi.org:10.1038/s41586-018-0616-y Piewngam P et al (2023) Probiotic for pathogen-specific Staphylococcus aureus decolonisation in Thailand: a phase 2, double-blind, randomised, placebo-controlled trial. Lancet Microbe 4:e75–e83. https://doi.org:10.1016/s2666-5247(22)00322-6 de Goffau MC, Charnock-Jones DS, Smith GCS, Parkhill J (2021) Batch effects account for the main findings of an in utero human intestinal bacterial colonization study. Microbiome 9:6. https://doi.org:10.1186/s40168-020-00949-z de Goffau MC et al (2019) Human placenta has no microbiome but can contain potential pathogens. Nature 572:329–334. https://doi.org:10.1038/s41586-019-1451-5 de Goffau MC et al (2018) Recognizing the reagent microbiome. Nat Microbiol 3:851–853. https://doi.org:10.1038/s41564-018-0202-y Frank DN et al (2010) The human nasal microbiota and Staphylococcus aureus carriage. PLoS ONE 5:e10598. https://doi.org:10.1371/journal.pone.0010598 Williams RE (1963) Healthy carriage of Staphylococcus aureus: its prevalence and importance. Bacteriological reviews 27:56–71 Harrison EM et al (2016) Validation of self-administered nasal swabs and postage for the isolation of Staphylococcus aureus. J Med Microbiol 65:1434–1437. https://doi.org:10.1099/jmm.0.000381 Verhoeven PO et al (2012) An algorithm based on one or two nasal samples is accurate to identify persistent nasal carriers of Staphylococcus aureus. Clin Microbiol infection: official publication Eur Soc Clin Microbiol Infect Dis 18:551–557. https://doi.org:10.1111/j.1469-0691.2011.03611.x Halablab MA, Hijazi SM, Fawzi MA, Araj GF (2010) Staphylococcus aureus nasal carriage rate and associated risk factors in individuals in the community. Epidemiol Infect 138:702–706. https://doi.org:10.1017/S0950268809991233 Andersen PS et al (2013) Risk factors for Staphylococcus aureus nasal colonization in Danish middle-aged and elderly twins. Eur J Clin Microbiol Infect diseases: official publication Eur Soc Clin Microbiol 32:1321–1326. https://doi.org:10.1007/s10096-013-1882-0 Andy Liaw MW randomForest: Breiman and Cutler's random forests for classification and regression , https://rdrr.io/rforge/randomForest/man/importance.html ( Richardson EJ et al (2018) Gene exchange drives the ecological success of a multi-host bacterial pathogen. Nat Ecol Evol 2:1468–1478. https://doi.org:10.1038/s41559-018-0617-0 Costello EK et al (2009) Bacterial community variation in human body habitats across space and time. Science 326:1694–1697. https://doi.org:10.1126/science.1177486 Ingham AC et al (2021) Dynamics of the Human Nasal Microbiota and Staphylococcus aureus CC398 Carriage in Pig Truck Drivers across One Workweek. Appl Environ Microbiol 87:e0122521. https://doi.org:10.1128/AEM.01225-21 Liu CM et al (2015) Staphylococcus aureus and the ecology of the nasal microbiome. Sci Adv 1. https://doi.org:10.1126/sciadv.1400216 Yan M et al (2013) Nasal microenvironments and interspecific interactions influence nasal microbiota complexity and S. aureus carriage. Cell Host Microbe 14. https://doi.org:10.1016/j.chom.2013.11.005 Tamkin E et al (2025) Airway Corynebacterium interfere with Streptococcus pneumoniae and Staphylococcus aureus infection and express secreted factors selectively targeting each pathogen. Infect Immun 93:e0044524. https://doi.org:10.1128/iai.00445-24 Huang S et al (2022) Corynebacterium accolens inhibits Staphylococcus aureus induced mucosal barrier disruption. Front Microbiol 13:984741. https://doi.org:10.3389/fmicb.2022.984741 de Goffau MC, Yang X, van Dijl JM, Harmsen HJ (2009) Bacterial pleomorphism and competition in a relative humidity gradient. Environ Microbiol 11:809–822. https://doi.org:10.1111/j.1462-2920.2008.01802.x Denis O (2017) Route of transmission of Staphylococcus aureus. Lancet Infect Dis 17:124–125. https://doi.org:10.1016/S1473-3099(16)30512-6 Coll F et al (2025) The mutational landscape of Staphylococcus aureus during colonisation. Nat Commun 16:302. https://doi.org:10.1038/s41467-024-55186-x Aggarwal D et al (2023) Optimization of high-throughput 16S rRNA gene amplicon sequencing: an assessment of PCR pooling, mastermix use and contamination. Microb Genom 9. https://doi.org:10.1099/mgen.0.001115 Understanding the biological basis of Staphylococcus aureus CARRIAGE. , https://www.carriagestudy.org.uk ( P S (2019) MiSeq SOP , https://mothur.org/wiki/miseq_sop/ Edgar RC, Haas BJ, Clemente JC, Quince C, Knight R (2011) UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 27:2194–2200. https://doi.org:10.1093/bioinformatics/btr381 Eren AM et al (2013) Differentiating between closely related microbial taxa using 16S rRNA gene data. Methods Ecol Evol 4. https://doi.org:10.1111/2041-210X.12114 . Oligotyping Ludwig W et al (2004) ARB: a software environment for sequence data. Nucleic Acids Res 32:1363–1371. https://doi.org:10.1093/nar/gkh293 Morgulis A et al (2008) Database indexing for production MegaBLAST searches. Bioinformatics 24:1757–1764. https://doi.org:10.1093/bioinformatics/btn322 McMurdie PJ, Holmes S (2013) phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE 8:e61217. https://doi.org:10.1371/journal.pone.0061217 Davis NM, Proctor DM, Holmes SP, Relman DA, Callahan BJ (2018) Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome 6:226. https://doi.org:10.1186/s40168-018-0605-2 Salter SJ et al (2014) Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol 12:87. https://doi.org:10.1186/s12915-014-0087-z de Goffau MC et al (2022) Gut microbiomes from Gambian infants reveal the development of a non-industrialized Prevotella-based trophic network. Nat Microbiol 7:132–144. https://doi.org:10.1038/s41564-021-01023-6 Oksanen J, Blanchet SG, Kindt F, Legendre R, Minchin P, O'Hara P, Solymos R, Stevens P, Szoecs M, Wagner E, Barbour H, Bedward M, Bolker M, Borcard B, Carvalho D, Chirico G, De Caceres M, Durand M, Evangelista S, FitzJohn H, Friendly R, Furneaux M, Hannigan B, Hill G, Lahti M, McGlinn L, Ouellette D, Ribeiro Cunha M, Smith E, Stier T, Ter Braak A, Weedon C (2022) J. _vegan: Community Ecology Package_. R package version 2.6-4 , . Lin H, Peddada SD (2020) Analysis of compositions of microbiomes with bias correction. Nat Commun 11:3514. https://doi.org:10.1038/s41467-020-17041-7 Mirzayi C et al (2021) Reporting guidelines for human microbiome research: the STORMS checklist. Nat Med 27:1885–1892. https://doi.org:10.1038/s41591-021-01552-x Wood DE, Lu J, Langmead B (2019) Improved metagenomic analysis with Kraken 2. Genome Biol 20:257. https://doi.org:10.1186/s13059-019-1891-0 Seemann T (2014) Prokka: rapid prokaryotic genome annotation. Bioinformatics 30:2068–2069. https://doi.org:10.1093/bioinformatics/btu153 Ewels P, Magnusson M, Lundin S, Kaller M (2016) MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32:3047–3048. https://doi.org:10.1093/bioinformatics/btw354 Tonkin-Hill G et al (2020) Producing polished prokaryotic pangenomes with the Panaroo pipeline. Genome Biol 21:180. https://doi.org:10.1186/s13059-020-02090-4 Jolley KA, Bray JE, Maiden MCJ (2018) Open-access bacterial population genomics: BIGSdb software, the PubMLST.org website and their applications. Wellcome Open Res 3:124. https://doi.org:10.12688/wellcomeopenres.14826.1 Page AJ et al (2016) SNP-sites: rapid efficient extraction of SNPs from multi-FASTA alignments. Microb Genom 2:e000056. https://doi.org:10.1099/mgen.0.000056 Minh BQ et al (2020) IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Mol Biol Evol 37:1530–1534. https://doi.org:10.1093/molbev/msaa015 Kalyaanamoorthy S, Minh BQ, Wong TKF, von Haeseler A, Jermiin LS (2017) ModelFinder: fast model selection for accurate phylogenetic estimates. Nat Methods 14:587–589. https://doi.org:10.1038/nmeth.4285 Minh BQ, Nguyen MA, von Haeseler A (2013) Ultrafast approximation for phylogenetic bootstrap. Mol Biol Evol 30:1188–1195. https://doi.org:10.1093/molbev/mst024 Additional Declarations Yes there is potential Competing Interest. A.B. reports institutional grants from AstraZeneca, Bayer, Biogen, BioMarin, Bioverativ, Novartis, Regeneron and Sanofi. J.D. serves on scientific advisory boards for AstraZeneca, Novartis, and UK Biobank, and has received multiple grants from academic, charitable and industry sources outside of the submitted work. Supplementary Files AggarwaletalSupplementaryMaterial210225.docx Supplementary Material TableS6.xlsx Table S6 TableS7.xlsx Table S7 TableS8.xls Table S8 TableS9.xls Table S9 Cite Share Download PDF Status: Published Journal Publication published 02 Dec, 2025 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6079410","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":447812027,"identity":"f4ba9ac7-1b26-4645-8575-43cf7632f2a7","order_by":0,"name":"Dinesh Aggarwal","email":"","orcid":"https://orcid.org/0000-0002-5938-8172","institution":"Imperial College","correspondingAuthor":false,"prefix":"","firstName":"Dinesh","middleName":"","lastName":"Aggarwal","suffix":""},{"id":447812028,"identity":"149edc94-2df8-4d2f-967a-1903bf457f08","order_by":1,"name":"Katherine L. Bellis","email":"","orcid":"https://orcid.org/0000-0001-5368-7027","institution":"Department of Medicine, University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Katherine","middleName":"L.","lastName":"Bellis","suffix":""},{"id":447812029,"identity":"2ef788cf-46ff-4161-ba99-0f1a476f38cf","order_by":2,"name":"Beth Blane","email":"","orcid":"","institution":"Department of Medicine, University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Beth","middleName":"","lastName":"Blane","suffix":""},{"id":447812030,"identity":"7ae5cfa0-c259-46e1-9974-c6a76ec9fba9","order_by":3,"name":"Marcus C. de Goffau","email":"","orcid":"","institution":"Tytgat Institute for Liver and Intestinal Research, Amsterdam University Medical Centers","correspondingAuthor":false,"prefix":"","firstName":"Marcus","middleName":"C.","lastName":"de Goffau","suffix":""},{"id":447812031,"identity":"ae187595-b458-43b4-9d39-161ef27d83f7","order_by":4,"name":"Josef Wagner","email":"","orcid":"","institution":"Wellcome Sanger Institue","correspondingAuthor":false,"prefix":"","firstName":"Josef","middleName":"","lastName":"Wagner","suffix":""},{"id":447812032,"identity":"506cdf7b-0f8a-4279-85db-40a5afa882fa","order_by":5,"name":"Duncan Y.K. Ng","email":"","orcid":"https://orcid.org/0000-0003-1561-8244","institution":"The Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Duncan","middleName":"Y.K.","lastName":"Ng","suffix":""},{"id":447812033,"identity":"ad722a93-899b-49b5-8024-72ebb6163cea","order_by":6,"name":"Kathy E. Raven","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Kathy","middleName":"E.","lastName":"Raven","suffix":""},{"id":447812034,"identity":"15daaa3b-97e7-4232-ae04-0c98aaa99ded","order_by":7,"name":"Plamena Naydenova","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Plamena","middleName":"","lastName":"Naydenova","suffix":""},{"id":447812035,"identity":"a5c69fbf-38b9-40d7-84cb-9a23c456469b","order_by":8,"name":"Stephen Kaptoge","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Stephen","middleName":"","lastName":"Kaptoge","suffix":""},{"id":447812036,"identity":"bff35ffc-b717-4d74-bc21-ae882b249e9b","order_by":9,"name":"Susan Burton","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Susan","middleName":"","lastName":"Burton","suffix":""},{"id":447812037,"identity":"43815fdf-25c7-4814-88c1-2de608d75893","order_by":10,"name":"Rachel Henry","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Rachel","middleName":"","lastName":"Henry","suffix":""},{"id":447812038,"identity":"bafd9082-8cdd-49e5-83b8-557fa2c7c4ad","order_by":11,"name":"Catherine Perry","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Catherine","middleName":"","lastName":"Perry","suffix":""},{"id":447812039,"identity":"491504f3-5533-457a-8710-aa946ead3014","order_by":12,"name":"Matthew R. Walker","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"R.","lastName":"Walker","suffix":""},{"id":447812040,"identity":"d1ec1a70-5f76-44fc-84f3-bde750346baa","order_by":13,"name":"Carmel Moore","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Carmel","middleName":"","lastName":"Moore","suffix":""},{"id":447812041,"identity":"efbda61f-0028-4f8a-adb3-182798df8c52","order_by":14,"name":"Carol Churcher","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Carol","middleName":"","lastName":"Churcher","suffix":""},{"id":447812042,"identity":"b96a120a-0849-4741-8471-bac9e7d7c714","order_by":15,"name":"Sophia T. Girgis","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Sophia","middleName":"T.","lastName":"Girgis","suffix":""},{"id":447812043,"identity":"d1abbbdc-bd5d-4309-97d1-cdc9aef3c638","order_by":16,"name":"Catarina Ribeiro de Sousa","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Catarina","middleName":"Ribeiro","lastName":"de Sousa","suffix":""},{"id":447812044,"identity":"08da595c-fde2-415e-9391-dafec93d85a0","order_by":17,"name":"Lauma Sarkane","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Lauma","middleName":"","lastName":"Sarkane","suffix":""},{"id":447812045,"identity":"ce74eff7-7f01-4f1c-95d5-26a2fde1f830","order_by":18,"name":"Joe Brennan","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Joe","middleName":"","lastName":"Brennan","suffix":""},{"id":447812046,"identity":"5e19799e-c822-4aff-92f1-03d2bec3aa26","order_by":19,"name":"Asha Akram","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Asha","middleName":"","lastName":"Akram","suffix":""},{"id":447812047,"identity":"26e8ddec-eff3-40ab-abb7-36462725e265","order_by":20,"name":"Shannon Duthie","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Shannon","middleName":"","lastName":"Duthie","suffix":""},{"id":447812048,"identity":"4bbf6f45-6dd8-4427-85e5-ccdca4c80343","order_by":21,"name":"Elisha Johnson","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Elisha","middleName":"","lastName":"Johnson","suffix":""},{"id":447812049,"identity":"4053b4c8-b054-4990-90b7-3896629bc1e4","order_by":22,"name":"Mercedesz Juhasz","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Mercedesz","middleName":"","lastName":"Juhasz","suffix":""},{"id":447812050,"identity":"8145392b-5476-4ee4-aa7d-8eb2040ab4f9","order_by":23,"name":"David Anderson","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Anderson","suffix":""},{"id":447812051,"identity":"477ac937-40d4-4450-a1cf-664516e4150f","order_by":24,"name":"Susan Irvine","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Susan","middleName":"","lastName":"Irvine","suffix":""},{"id":447812052,"identity":"9cd43ee2-3f36-40d7-84aa-b0b84523ca18","order_by":25,"name":"Amy McMahon","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Amy","middleName":"","lastName":"McMahon","suffix":""},{"id":447812053,"identity":"0eb3c425-4404-4ff4-8994-55dedf6acbcf","order_by":26,"name":"Liz Lay","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Liz","middleName":"","lastName":"Lay","suffix":""},{"id":447812054,"identity":"e18824fe-b607-4a68-86e0-4602fd78eab2","order_by":27,"name":"Claire Raisen","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Claire","middleName":"","lastName":"Raisen","suffix":""},{"id":447812055,"identity":"74c07d0a-4828-456f-9217-913860f9337f","order_by":28,"name":"Xiaoliang Ba","email":"","orcid":"https://orcid.org/0000-0002-3882-3585","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Xiaoliang","middleName":"","lastName":"Ba","suffix":""},{"id":447812056,"identity":"a30d7cd8-e1e5-42cf-8672-ddd64149314f","order_by":29,"name":"Mark Holmes","email":"","orcid":"https://orcid.org/0000-0002-5454-1625","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Mark","middleName":"","lastName":"Holmes","suffix":""},{"id":447812057,"identity":"45a7fdb3-af41-4447-857f-ac291468fc74","order_by":30,"name":"Andries van Tonder","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Andries","middleName":"van","lastName":"Tonder","suffix":""},{"id":447812058,"identity":"bcf74730-1fdd-4749-aa8d-ba7953c53b83","order_by":31,"name":"Emanuele Di Angelantonio","email":"","orcid":"https://orcid.org/0000-0001-8776-6719","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Emanuele","middleName":"Di","lastName":"Angelantonio","suffix":""},{"id":447812059,"identity":"98931f6e-05a1-4eb8-886a-33a8fef33275","order_by":32,"name":"Adam Butterworth","email":"","orcid":"https://orcid.org/0000-0002-6915-9015","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Adam","middleName":"","lastName":"Butterworth","suffix":""},{"id":447812060,"identity":"bf553b45-c9ba-4de5-97c8-ef2c73bcc0c3","order_by":33,"name":"Joan A. Geoghegan","email":"","orcid":"https://orcid.org/0000-0002-3788-0668","institution":"University of Birmingham","correspondingAuthor":false,"prefix":"","firstName":"Joan","middleName":"A.","lastName":"Geoghegan","suffix":""},{"id":447812061,"identity":"7ad7d8e0-89aa-4bff-8bc3-1cd80b6311d8","order_by":34,"name":"John Danesh","email":"","orcid":"","institution":"Unveristy of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Danesh","suffix":""},{"id":447812062,"identity":"2a9e6fd0-16a1-49a2-affe-56d967119b36","order_by":35,"name":"Julian Parkhill","email":"","orcid":"https://orcid.org/0000-0002-7069-5958","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Julian","middleName":"","lastName":"Parkhill","suffix":""},{"id":447812063,"identity":"ac51b86a-54f0-4688-9462-5dfeddbd5ffd","order_by":36,"name":"Sharon J. Peacock","email":"","orcid":"https://orcid.org/0000-0002-1718-2782","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Sharon","middleName":"J.","lastName":"Peacock","suffix":""},{"id":447812026,"identity":"9eed8a37-90a3-41e9-a053-791da11d6fc8","order_by":37,"name":"Ewan M. Harrison","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBElEQVRIiWNgGAWjYDACCYbkBx8qgIwDDGwMDxDiCfi0PDOccQaqBUkdPi2MD6R520jRwi/dnGDMO2+bHN/5w88eJDDYyes2MD/8wNiWhlOL5JxjCQ/nbrttLHkjzdwggSHZcNsBNmMJxrYcnFoMbuQkGLzddjtxww0eNokEhgOM2w4wmDEwtlXg1GJ/I/+DBO+c2/Ubzp8Ba7HfdoD9G14tBhIJCZK8DbcTDA7kgLUkbjvAA7IFt8MkbiSkGc44dttw5o00M4kEg+TkbYd5iiUSzuH2Pv+MBGBU1tyWB4WYxIcKO9ttx9s3fvhQloxTC7o7gZiZAW9EjoJRMApGwSggAgAAIHhaRrvN7YoAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0003-2720-0507","institution":"Wellcome Sanger Institue","correspondingAuthor":true,"prefix":"","firstName":"Ewan","middleName":"M.","lastName":"Harrison","suffix":""}],"badges":[],"createdAt":"2025-02-21 12:20:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6079410/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6079410/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-025-66564-4","type":"published","date":"2025-12-02T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":81616910,"identity":"ded2c53b-aeea-44e4-a7fc-8b9cd9cae243","added_by":"auto","created_at":"2025-04-29 08:24:42","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":762993,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy design and nasal diversity and composition by \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eStaphylococcus aureus\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e colonisation status and \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eStaphylococcus aureus\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e culture result.\u003c/strong\u003e (a) illustration of study design and cohort (b-e) Box plots comparing Alpha diversity (Shannon and Simpson) from nasal samples by (b,c) \u003cem\u003eStaphylococcus aureus\u003c/em\u003e colonisation status and (d,e) \u003cem\u003eStaphylococcus aureus\u003c/em\u003eculture result. (f-i) Ordination plots representing Beta diversity by Bray-Curtis distance and coloured by (f,g) \u003cem\u003eStaphylococcus aureus\u003c/em\u003e colonisation status and (h,i) \u003cem\u003eStaphylococcus aureus\u003c/em\u003e culture result. (f,h) show NMDS plots (g,i) show PCoA plots. (f,g) Bray-Curtis distance between colonisation states significantly differed by PERMANOVA analysis (\u003cem\u003eF\u003c/em\u003e(2) = 36.67, p \u0026lt; 0.001). (h,i) Bray-Curtis distance between \u003cem\u003eStaphylococcus aureus\u003c/em\u003eculture positive and negative samples significantly differed by PERMANOVA analysis (\u003cem\u003eF\u003c/em\u003e(2) = 59.01, p \u0026lt; 0.001). Data ellipses represent the 95% confidence level that values lie within this space, assuming a multivariate t-distribution. (j) Mean microbial composition of all samples at a species level (k) Mean microbial composition of samples by \u003cem\u003eS. aureus\u003c/em\u003ecolonisation status at a species level (l) Microbial composition of samples across the study dataset, separated by colonisation status at a species level (m) Microbial composition of samples across the study dataset, separated by \u003cem\u003eS. aureus\u003c/em\u003e nasal swab culture result at a species level. (j-m) Top 17 species represented. (l,m) Samples sorted by Bray-Curtis similarity.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/f2e620bb0d8956d5ba1f726c.png"},{"id":81616883,"identity":"f975af04-1d08-4bfe-811a-32320bfdd999","added_by":"auto","created_at":"2025-04-29 08:24:25","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":559614,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMicrobial community state types and differential abundance of species observed in the anterior nares. \u003c/strong\u003e(a) Heatmap of species abundances in CARRIAGE nasal samples. Samples are ordered by hierarchical clustering using Bray-Curtis distances based on the compositional, relative abundance data, represented by the dendrogram. Prevalence of each species across the samples is represented by the horizontal bar plots. Community state types (CSTs) and \u003cem\u003eS. aureus\u003c/em\u003e colonisation status of samples are represented above the heatmap. Seven distinct CSTs were identified from the selection of hierarchical clusters determined by calculating a gap statistic on the Bray-Curtis distance. (b) Bacterial species dominating each CST. (c) Composition of CSTs by colonisation status. (d) Composition of \u003cem\u003eS. aureus\u003c/em\u003e colonisation status by CSTs. (e) Composition of each CST by sex. *significant difference (p\u0026lt;0.05) by sex in the multinomial logistic regression model p\u0026lt;0.05 (f) Differential abundance of species by nasal colonisation status using ANCOM-BC2. Log-fold (natural log) changes as compared to \u003cem\u003eS. aureus\u003c/em\u003e non-carriers. Column one compares persistent carriers against non-carriers and column two compares intermittent carriers against non-carriers.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/b73aa4283370e2a77ce691ba.png"},{"id":81617121,"identity":"363b52fb-eb40-4315-b03f-840d3a86541e","added_by":"auto","created_at":"2025-04-29 08:32:25","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":644083,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMicrobial composition of species in the anterior nares by the number of positive\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003e S. aureus\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e swabs, with a focus on intermittent carriers. \u003c/strong\u003e(a-d) Ordination plots representing Beta diversity by Bray-Curtis distance and coloured by \u003cem\u003eStaphylococcus aureus\u003c/em\u003e colonisation status and the number of \u003cem\u003eS. aureus \u003c/em\u003eculture-positive swabs relating to each participant represented. (a, c) show NMDS plots (b, d) show PCoA plots. (a-b) Figure 1 panels f to g have been reproduced but with \u003cem\u003eS. aureus \u003c/em\u003enon-carriers and persistent carriers faded into the background. The values representing intermittent carriers on the ordination plots of Bray-Curtis distance visibly span both the non-carrier and persistent carrier clusters. (c-d) These plots represent the same Bray-Curtis distances as shown on panel a to b but with points coloured by the number of positive swabs. Despite limited numbers, it is apparent that there is greater overlap of the non-carriers (0 positive swabs) with the participants with 1 positive swab individuals, and a similar relationship is seen between the participants with 2 positive swabs and the persistent carriers (3 positive swabs). (e) mean abundance by the number of swabs positive for \u003cem\u003eS. aureus \u003c/em\u003eincluding non-, intermittent and persistent carriers (f) Microbial composition represented by relative abundance of species residing in the anterior nares of individual intermittent carriers, comparing the number of \u003cem\u003eStaphylococcus aureus\u003c/em\u003e culture positive swabs obtained (one vs two). (g) Microbial composition of the anterior nares from intermittent carriers represented as a heatmap. Heatmap of species abundances in nasal samples from intermittent carriers. Samples are ordered by hierarchical clustering using Bray-Curtis distances on the compositional relative abundance data. Prevalence of each species is highlighted in the horizontal bar plots. The dashed red line represents splitting of hierarchical clustering dendrogram in seven community state types, as determined by the gap statistic. Participants with two positive swabs appear to have a higher abundance of \u003cem\u003eS. aureus\u003c/em\u003e.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/12d96f79fcef20bb855e7271.png"},{"id":81616885,"identity":"a9730a0e-9503-4e78-a5b6-1b2aea448a07","added_by":"auto","created_at":"2025-04-29 08:24:25","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":76928,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRandom forest classifier of the nasal microbiome data. \u003c/strong\u003ea) ROC curves demonstrating model performance for classification of non-carriers vs others (grey line), persistent carriers vs others (blue line), and intermittent carriers vs others (red line). The multi-class area under the curve was calculated as 76\u003cem\u003e.\u003c/em\u003e8%. (b) Performance of the random forest model to predict the nasal microbiome. Values provided as percentages. (c) Feature importance as determined by mean decrease in gini index from the random forest classifier (d) Feature importance as determined by mean decrease in model accuracy from the random forest classifier.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/774fe6182b1bc64dbead5b16.png"},{"id":81616887,"identity":"09e6b7f6-235c-45f0-968d-ac241e64b36a","added_by":"auto","created_at":"2025-04-29 08:24:25","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":416580,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVariation of the anterior nares microbiome with the \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eStaphylococcus aureus\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003ephylogeny. \u003c/strong\u003e(a) Maximum-likelihood tree of \u003cem\u003eS. aureus\u003c/em\u003e whole-genome sequences cultured from persistent and intermittent carriers labelled with their associated carriage status, sequence-type and microbiome. (b) Box plots comparing rarified \u003cem\u003eS. aureus\u003c/em\u003e abundance by cluster (c,d) Ordination plots representing Beta diversity by Bray-Curtis distance and coloured by the phylogenetic clusters (A or B) representing the bifurcation of the tree. Bray-Curtis distance between samples representing phylogenetic clusters A and B samples significantly differed, although weakly, by PERMANOVA analysis (\u003cem\u003eF\u003c/em\u003e(2) = 2.33, p = 0.04). (c) shows an NMDS plot (d) shows a PCoA plot. Data ellipses represent the 95% confidence level that values lie within this space assuming a multivariate t-distribution. (e) Graphical representation of abstract.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/8832b2beb0cfeef00caaf4fc.png"},{"id":97322714,"identity":"9db67d58-ae14-4a03-aa6d-64b731a5c775","added_by":"auto","created_at":"2025-12-03 08:09:26","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3993177,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/41ba8d98-6dba-4e3e-aa1b-0e68448ba7a9.pdf"},{"id":81617127,"identity":"5114df0c-348d-4ee0-9ea2-7b51727af2cf","added_by":"auto","created_at":"2025-04-29 08:32:25","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":9859454,"visible":true,"origin":"","legend":"Supplementary Material","description":"","filename":"AggarwaletalSupplementaryMaterial210225.docx","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/4f2bd7d9fb94c537e9ed936c.docx"},{"id":81618473,"identity":"0fd9217b-53a9-4610-b963-8a609d147365","added_by":"auto","created_at":"2025-04-29 08:48:25","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":83304,"visible":true,"origin":"","legend":"Table S6","description":"","filename":"TableS6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/76bfbce36cddada2f1a16e08.xlsx"},{"id":81617119,"identity":"d12e5f55-2945-4ba9-b089-3363eecc24b0","added_by":"auto","created_at":"2025-04-29 08:32:25","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":18243,"visible":true,"origin":"","legend":"\u003cp\u003eTable S7\u003c/p\u003e","description":"","filename":"TableS7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/d8fb280fdad758c1be6aae64.xlsx"},{"id":81616902,"identity":"0ead11d2-589c-40de-8fa8-d9eb4a6f888d","added_by":"auto","created_at":"2025-04-29 08:24:25","extension":"xls","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":3016192,"visible":true,"origin":"","legend":"Table S8","description":"","filename":"TableS8.xls","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/f28e4a2e7d1ccb4786c32883.xls"},{"id":81619737,"identity":"1658788e-71ff-4252-b415-00fc5219dd59","added_by":"auto","created_at":"2025-04-29 08:56:25","extension":"xls","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":356864,"visible":true,"origin":"","legend":"Table S9","description":"","filename":"TableS9.xls","url":"https://assets-eu.researchsquare.com/files/rs-6079410/v1/da49e29b6edc654492a314ee.xls"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nA.B. reports institutional grants from AstraZeneca, Bayer, Biogen, BioMarin, Bioverativ, Novartis, Regeneron and Sanofi. J.D. serves on scientific advisory boards for AstraZeneca, Novartis, and UK Biobank, and has received multiple grants from academic, charitable and industry sources outside of the submitted work.","formattedTitle":"The nasal microbiome redefines Staphylococcus aureus colonisation","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe human nose is populated by a range of bacterial species which constitute the nasal microbiota, including the important commensal and opportunistic pathogen \u003cem\u003eStaphylococcus aureus\u003c/em\u003e \u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eS. aureus\u003c/em\u003e nasal carriage is clinically important; carriers are at greater risk of \u003cem\u003eS. aureus\u003c/em\u003e infection, often caused by the colonising strain \u003csup\u003e\u003cspan additionalcitationids=\"CR5\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e whilst decolonisation can reduce infection rates \u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Based on longitudinal sampling, \u003cem\u003eS. aureus\u003c/em\u003e nasal colonisation states have historically been divided into persistent, intermittent, and non-carrier \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. However, it has been hypothesised that there may be only two categories (persistent carriers and non-carriers); as compared to intermittent carriers persistent carriers: (a) have higher \u003cem\u003eS. aureus\u003c/em\u003e loads \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e (b) are more likely to become recolonised and maintain colonisation in experimental colonisation, (c) will preferentially select their own strain when inoculated with multiple strains \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, and (d) have higher antibody levels to a limited number of \u003cem\u003eS. aureus\u003c/em\u003e antigens (in small-scale studies) \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eMultiple host factors have been identified that influence \u003cem\u003eS. aureus\u003c/em\u003e carriage, and colonisation is higher in adult males and children \u003csup\u003e\u003cspan additionalcitationids=\"CR12\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, though many of these studies are small and in unrepresentative cohorts \u003csup\u003e\u003cspan additionalcitationids=\"CR13\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Interactions between the microbial residents of the anterior nares have been described between \u003cem\u003eS. aureus\u003c/em\u003e and other members of the nasal microbiome \u003csup\u003e\u003cspan additionalcitationids=\"CR16 CR17\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. For example, both \u003cem\u003eStaphylococcus epidermidis\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e and \u003cem\u003eStaphylococcus lugdunensis\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e have been shown to produce distinct compounds that inhibit \u003cem\u003eS. aureus\u003c/em\u003e growth.\u003c/p\u003e \u003cp\u003eUnlike the gut microbiome, only a limited number of small studies have investigated the nasal microbiome. Yan \u003cem\u003eet al\u003c/em\u003e. examined the nasal microbiome of 12 individuals at the anterior nares and two sites in the inner part of the nasal cavity, revealing similarity in the dominant species between sites, but a lower overall diversity in the anterior nares \u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e. Analysis of the microbiome composition of culture-defined persistent or non-\u003cem\u003eS. aureus\u003c/em\u003e carriers revealed that \u003cem\u003eS. aureus\u003c/em\u003e had both an antagonistic relationship with \u003cem\u003eCorynebacterium pseudodiptheriticum\u003c/em\u003e and synergism with \u003cem\u003eCorynebacterium accolens\u003c/em\u003e, that was confirmed experimentally \u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e. A larger study of eighty-six older (mean age\u0026thinsp;~\u0026thinsp;65) twin pairs defined seven community state types (CST; distinct groups of bacteria) in the nose, identified negative associations of \u003cem\u003eS. aureus\u003c/em\u003e with \u003cem\u003eDolosigranulum\u003c/em\u003e, \u003cem\u003eSimonsiella\u003c/em\u003e, \u003cem\u003ePropionibacterium granulosum;\u003c/em\u003e a positive association with \u003cem\u003eS. epidermidis\u003c/em\u003e, and found a lower overall bacterial density and \u003cem\u003eS. aureus\u003c/em\u003e 16S gene copy abundance amongst women \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. In early life, the assembly of the nasal microbiome is weakly influenced by the maternal microbiome, while environmental exposures such as daycare have a greater impact \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, both the type of birth delivery and breast feeding influence the infant nasal microbiome with some variation between studies (reviewed in \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e). It is clear that \u003cem\u003eS. aureus\u003c/em\u003e colonises the nasopharynx in the first weeks of life \u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, and several species have been reported to support \u003cem\u003eS. aureus\u003c/em\u003e colonisation in infants \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, and be inversely correlated with maternal \u003cem\u003eDolosigranulum pigrum\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. More recent metanalysis of paediatric (n\u0026thinsp;=\u0026thinsp;99 individuals) and adult (n\u0026thinsp;=\u0026thinsp;210) 16s rRNA gene-based studies revealed positive integrations between \u003cem\u003eD. pigrum\u003c/em\u003e and both \u003cem\u003eC. pseudodiptheriticum\u003c/em\u003e, and \u003cem\u003eMoraxella nonliquefaciens\u003c/em\u003e in children \u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e. This trinity of species have been reported to be associated with greater stability of the nasal microbiome in early life \u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e. Similarly, in adults, four different \u003cem\u003eCorynebacterium\u003c/em\u003e species have been found positively associated with \u003cem\u003eD. pigrum\u003c/em\u003e, which in turn was negatively associated with \u003cem\u003eS. aureus\u003c/em\u003e, an association that was further demonstrated \u003cem\u003ein vitro\u003c/em\u003e \u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe use of antagonistic bacterial strains as live biotherapeutics (probiotics) is an attractive option to reduce \u003cem\u003eS. aureus\u003c/em\u003e nasal colonisation without the need for antibiotics. This concept was demonstrated using a \u003cem\u003eCorynebacterium\u003c/em\u003e sp. to successfully eradicate \u003cem\u003eS. aureus\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. In more recent work, \u003cem\u003eS. aureus\u003c/em\u003e was reported to be excluded from the gut in the presence of \u003cem\u003eB. subtilis\u003c/em\u003e via inhibition of pathogen signalling \u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, a finding that was translated into a clinical trial of \u003cem\u003eB. subtilis\u003c/em\u003e as a live biotherapeutic, which was successful eliminating viable \u003cem\u003eS. aureus\u003c/em\u003e from the gut and reducing but not eradicating the bacterial loads in the nose \u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn summary, while interactions between different bacterial species in the nasal microbiota have been identified including with \u003cem\u003eS. aureus\u003c/em\u003e, nasal microbiome studies have only involved small sample sizes often in selected populations which likely reduce the generalisability of the findings. In addition, in recent years the importance of systematic removal of contamination in microbiome studies, particularly lower biomass/complexity environment such the nasal microbiome, has been established \u003csup\u003e\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. This means studies (particularly those with small sample sizes) that have not systematically removed contamination risk being confounded. Critically, for understanding the nasal microbiome in relation to \u003cem\u003eS. aureus\u003c/em\u003e colonisation, no microbiome study greater than forty individuals \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e has yet included \u003cem\u003eS. aureus\u003c/em\u003e colonisation status as defined by longitudinal sampling and culture, which has been used to understand \u003cem\u003eS. aureus\u003c/em\u003e colonisation for 70 years \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHere, we utilise microbiome data from nasal swabs of 1,180 generally healthy community participants from across England in the CARRIAGE study, along with three weekly nasal swabs cultured for \u003cem\u003eS. aureus\u003c/em\u003e to determine the microbiome structure associated with nasal \u003cem\u003eS. aureus\u003c/em\u003e carriage, including evaluation of the validity of the current defined \u003cem\u003eS. aureus\u003c/em\u003e colonisation states (persistent, intermittent and non-carriers).\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eDetermination of the nasal microbiome in a large cohort\u003c/h2\u003e \u003cp\u003eTo study the biological basis of \u003cem\u003eS. aureus\u003c/em\u003e colonisation we sampled generally healthy adult blood donors from across England with three self-taken nasal swabs taken at weekly intervals (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). \u003cem\u003eS. aureus\u003c/em\u003e colonisation status was assessed by culture, and was defined as: (i) persistent colonisation (306/1091 (28.0%); three \u003cem\u003eS. aureus\u003c/em\u003e positive weekly nasal swabs, (ii) intermittent colonisation 191/1091 (17.5%); one or two positive swabs, and (iii) non-carrier 594/1091 (54.4%): no positive swabs, based on previous studies \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e (89 failed to return all samples). Participants had a mean age of 51.4 (median, 53) and 52.8% were female. A total of 1,756 samples, which included the first swabs of 1,180 participants underwent 16S rRNA gene sequencing to determine the microbiome composition (\u003cb\u003eSupplementary Fig. S1\u003c/b\u003e). After QC (see Methods and \u003cb\u003eSupplementary Table S2\u003c/b\u003e), 1,055 samples remained, and after rarefaction and a systematic analysis to remove any likely contaminants, 53 Operational Taxonomic Units (OTUs) (24 species level taxa) remained.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDifferences in within-sample microbiome diversity\u003c/h3\u003e\n\u003cp\u003eWe first investigated variation in Alpha diversity (measures of within-sample diversity) by \u003cem\u003eS. aureus\u003c/em\u003e culture-defined \u003cem\u003eS. aureus\u003c/em\u003e colonisation status, to determine differences in the microbiome between \u003cem\u003eS. aureus\u003c/em\u003e colonisation states. We found Alpha diversity was significantly lower in samples from persistent carriers when compared to non-carriers or intermittent carriers when using either the Shannon or Simpsons diversity metrics (both p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and found no significant difference between non-carriers and intermittent carriers using either (both p\u0026thinsp;\u0026gt;\u0026thinsp;0.2) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB, C). We observed significantly lower Alpha diversity in \u003cem\u003eS. aureus\u003c/em\u003e culture-positive samples than culture-negative samples when assessed with the Shannon diversity (p\u0026thinsp;=\u0026thinsp;0.047) but not with the Simpson diversity (p\u0026thinsp;=\u0026thinsp;0.09) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD, E).\u003c/p\u003e \u003cp\u003eWe next investigated Beta diversity (similarity or dissimilarity between two samples) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF-I) using Bray-curtis distance by colonisation status, which differed significantly (PERMANOVA analysis, \u003cem\u003eF\u003c/em\u003e(2)\u0026thinsp;=\u0026thinsp;36.67, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Distinct separation of samples by colonisation status could be clearly observed by non-metric multidimensional scaling (NMDS) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF) and PCoA (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eJ) ordination plots. However, samples from intermittent carriers did not form a distinct cluster, and instead overlapped within the persistent or the non-carrier clusters, but with more samples from intermittent carriers being clustered with the non-carriers as visualised by the overlap in data ellipses in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF, G. This suggests that the microbiomes of intermittent (or rather occasionally \u003cem\u003eS. aureus\u003c/em\u003e culture-positive) carriers are not distinct but typically more similar to non-carriers, with smaller numbers that have similar microbiomes to persistent carriers. Likewise, we observed similar distinct clusters between \u003cem\u003eS. aureus\u003c/em\u003e culture-positive and culture-negative samples on the ordination plots (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eH-I), as would be expected when \u003cem\u003eS. aureus\u003c/em\u003e culture positivity is associated with persistent colonisation. Again, the two groups defined by \u003cem\u003eS. aureus\u003c/em\u003e culture result differed significantly by PERMANOVA analysis (\u003cem\u003eF\u003c/em\u003e(2)\u0026thinsp;=\u0026thinsp;59.01, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eH, I). We only observed an association of sex with variation in the Bray-Curtis distance, as females (115/543, 21.2%) are less commonly persistent carriers compared to males (160/511, 31.3%) (p\u0026thinsp;=\u0026thinsp;\u0026lt;\u0026thinsp;0.001), but with a low F statistic and R\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e values (\u003cem\u003eF\u003c/em\u003e(2)\u0026thinsp;=\u0026thinsp;2.83, p\u0026thinsp;=\u0026thinsp;0.006, R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.39%). There was no association with sex, smoking, pet ownership, healthcare worker, chronic skin condition, and diabetes.\u003c/p\u003e\n\u003ch3\u003eCompositional differences by colonisation status and defining community state types\u003c/h3\u003e\n\u003cp\u003eTo visualise the causes of differences observed in Alpha and Beta diversity, we analysed species composition by \u003cem\u003eS. aureus\u003c/em\u003e colonisation status (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eJ-K). The lower Alpha diversity of the persistent carriers was associated with the dominance of \u003cem\u003eS. aureus\u003c/em\u003e in the species composition of this groups, compared to the \u0026lsquo;intermittent\u0026rsquo; and non-carriers. In contrast, the nasal microbiome of non-carriers is largely dominated by multiple \u003cem\u003eCorynebacterium\u003c/em\u003e species and \u003cem\u003eD. pigrum\u003c/em\u003e. We next examined species composition at the level of each participant\u0026rsquo;s sample, separated by colonisation state (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eL). This showed that amongst the 275 persistently colonised participants, \u003cem\u003eS. aureus\u003c/em\u003e was the dominant organism (\u0026gt;\u0026thinsp;50% of reads) for 136/275 (49.5%), and in a subset of 96/275 (34.9%) participants, \u003cem\u003eS. aureus\u003c/em\u003e represented\u0026thinsp;\u0026gt;\u0026thinsp;75% of reads. In comparison, \u0026gt;\u0026thinsp;50% of reads from \u003cem\u003eS. aureus\u003c/em\u003e was only seen in the 22/532 (4.1%), of \u003cem\u003eS. aureus\u003c/em\u003e culture-negative (non-carriers) and 26/169 (15.4%) occasionally \u003cem\u003eS. aureus\u003c/em\u003e culture-positive individuals (intermittent carriers). Instead, the non-carriers and subset of intermittent carriers were clearly dominated by three different \u003cem\u003eCorynebacterium\u003c/em\u003e species (\u003cem\u003eC. pseudodipthericum, C. jeikeium, and C. accolens\u003c/em\u003e) at abundances not seen in \u003cem\u003eS. aureus\u003c/em\u003e persistent carriers (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eL). Classifying individual samples by \u003cem\u003eS. aureus\u003c/em\u003e culture result revealed that \u003cem\u003eS. aureus\u003c/em\u003e was the predominant species (\u0026gt;\u0026thinsp;50% of reads) in 164/382 (42.9%) of the \u003cem\u003eS. aureus\u003c/em\u003e culture-positive samples, and only 32/672 (4.76%) of culture-negative samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eM). Men have higher \u003cem\u003eS. aureus\u003c/em\u003e culture positive rates \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e so we examined the possibility of bias in culture by sex; however, we did not find that culture-negative samples with a higher \u003cem\u003eS. aureus\u003c/em\u003e abundance (\u0026gt;\u0026thinsp;50% of reads) were more prevalent amongst females (17/32, 53.1%) compared to males (15/32, 46.9%).\u003c/p\u003e\n\u003ch3\u003eCommunity state types\u003c/h3\u003e\n\u003cp\u003eNext, we generated a heatmap of taxa abundance (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA), organised by hierarchical clustering by Bray-Curtis distance to examine the relationships between microbial residents of the anterior nares. We used this to define community state types (CSTs), i.e. samples with similar abundances of species which cluster together. To determine the number of clusters in the data, we calculated a gap statistic with ordination values using Bray-Curtis distances (\u003cb\u003eSupplementary Fig. S6\u003c/b\u003e). A total of 7 clusters were defined; we identified CSTs from the heatmap plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). From the heatmap, it is evident that individuals always \u003cem\u003eS. aureus\u003c/em\u003e culture-positive (persistent carriers) cluster to form the majority of CST1 (72.4%, 155/214), whilst those always \u003cem\u003eS. aureus\u003c/em\u003e culture-negative (non-carriers) are represented largely by the remaining CSTs (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB-C). Intermittent carriers are dispersed across the CSTs. Using a multinomial logistic regression model, we found men had a reduced relative risk for association with CST6 (O.R.= 0.53, 95% C.I.= 0.30\u0026ndash;0.93, p\u0026thinsp;=\u0026thinsp;\u0026lt;\u0026thinsp;0.05) and CST7 (O.R.=0.67, 95% C.I.= 0.47\u0026ndash;0.96, p\u0026thinsp;=\u0026thinsp;\u0026lt;\u0026thinsp;0.05) compared with CST1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE). No other significant associations with CSTs were observed. Adjusted odd-ratios are provided in \u003cb\u003eSupplementary Table S5\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe then formally evaluated differences in species abundances by colonisation status using ANCOM-BC2, which minimises the false discovery rate, using the unadjusted read count table. This demonstrated a significant differential abundance when comparing always \u003cem\u003eS. aureus\u003c/em\u003e culture-negative individuals (non-carriers) and always culture-positive individuals (persistent carriers) carriers, with a positive association of \u003cem\u003eS. aureus\u003c/em\u003e seen with persistent carriage (as expected), and a negative association seen with multiple \u003cem\u003eCorynebacterium\u003c/em\u003e species, \u003cem\u003eD. pigrum\u003c/em\u003e, \u003cem\u003eS. epidermidis\u003c/em\u003e, and \u003cem\u003eM. catarrhalis\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF \u003cb\u003eand Supplementary Table S4\u003c/b\u003e). No significant differences in species abundance other than \u003cem\u003eS. aureus\u003c/em\u003e between non-carriers and intermittent carriers is observed (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF \u003cb\u003eand Supplementary Table S4\u003c/b\u003e). Notably, persistent carriers had a greater log-fold change in \u003cem\u003eS. aureus\u003c/em\u003e when compared with occasionally \u003cem\u003eS. aureus\u003c/em\u003e positive individuals (intermittent carriers), in comparison to non-carriers, suggesting the relative abundance of \u003cem\u003eS. aureus\u003c/em\u003e may be driving its longitudinal carriage.\u003c/p\u003e \u003cp\u003eIn a subgroup of 34 participants, from the rarefied dataset to 10,000 reads, two or three samples (n\u0026thinsp;=\u0026thinsp;75) were available over consecutive weeks (\u003cb\u003eSupplementary Fig. S7-10\u003c/b\u003e). These included 13 persistent carriers, 7 intermittent carriers, and 14 non-carrier\u0026rsquo;. We examined the stability of the community in the anterior nares and correlated pairwise Alpha diversity of participants by colonisation status. Persistent (Spearman\u0026rsquo;s rho\u0026thinsp;=\u0026thinsp;0.54, p\u0026thinsp;=\u0026thinsp;0.028) and intermittent carriers (Spearman\u0026rsquo;s rho\u0026thinsp;=\u0026thinsp;0.79, p\u0026thinsp;=\u0026thinsp;0.028) were found to have greater stability compared to non-carriers (Spearman\u0026rsquo;s rho\u0026thinsp;=\u0026thinsp;0.30, p\u0026thinsp;=\u0026thinsp;0.268).\u003c/p\u003e\n\u003ch3\u003eFurther examining the microbiome of ‘intermittent’ carriage\u003c/h3\u003e\n\u003cp\u003eHaving observed that the majority of intermittent carriers clustered with \u003cem\u003eS. aureus\u003c/em\u003e non-carriers group (e.g. overlapping data ellipses in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF, G), we hypothesised that intermittent carriers could be misclassified non- or persistent carriers. We examined differences in Alpha diversity between the one and two swab positive intermittent subgroup (\u003cb\u003eSupplementary Fig. S11\u003c/b\u003e), using only samples with greater than 10,000 reads. We found no significant difference in Alpha diversity when comparing samples with one \u003cem\u003eS. aureus\u003c/em\u003e positive swab compared with two (p\u0026thinsp;=\u0026thinsp;0.21). Beta diversity by Bray-Curtis index between samples with one or two positive \u003cem\u003eS. aureus\u003c/em\u003e swabs did differ significantly by PERMANOVA analysis (\u003cem\u003eF\u003c/em\u003e(2)\u0026thinsp;=\u0026thinsp;3.19, p\u0026thinsp;=\u0026thinsp;0.003), suggesting that these groups have differing microbial compositions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-D).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe next explored the abundance of species across the samples depending on the number of swabs which were positive for \u003cem\u003eS. aureus\u003c/em\u003e. There is a clear continuous trend in the variation in abundance from zero to three positive swabs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF). We then subset the participants representing intermittent carriers (n\u0026thinsp;=\u0026thinsp;169) from the dataset to examine if these were two distinct populations (rather than one) based on the number of \u003cem\u003eS. aureus\u003c/em\u003e positive swabs (one swab, n\u0026thinsp;=\u0026thinsp;103 and two swabs, n\u0026thinsp;=\u0026thinsp;66). From examination of the species composition of individual samples, a different microbial community structure is apparent for intermittent carriers who are positive for two swabs compared to those with one swab (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF).\u003c/p\u003e \u003cp\u003eWe formally analysed the differences in community structure using a heatmap of abundances from the samples of intermittent carriers, which displays a similar structure of clustering to that observed when comparing persistent carriers (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG). Again, we calculated a gap statistic, giving an optimal number of CSTs of 7 (same as full dataset), and the hierarchical clustered dendrogram was split accordingly (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG). On this heatmap, it is clear that the CSTs that are dominated by \u003cem\u003eCorynebacterium\u003c/em\u003e species, \u003cem\u003eD. pigrum\u003c/em\u003e and \u003cem\u003eS. epidermidis\u003c/em\u003e are associated with samples where participants had one positive \u003cem\u003eS. aureus\u003c/em\u003e culture. 18/66 (27.3%) intermittent carriers who had two positive \u003cem\u003eS. aureus\u003c/em\u003e cultures were associated with a CST dominated by \u003cem\u003eS. aureus\u003c/em\u003e, compared to 7/103 (6.8%) with one positive swab. These findings reflect similar observations for persistent carriers and non-carriers, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF, G), and provide further evidence that \u0026lsquo;intermittent carriers\u0026rsquo; are just misclassified persistent carriers or non-carriers.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003ePredicting colonisation status from the nasal microbiome\u003c/h2\u003e \u003cp\u003eWe next used a random forest model to establish whether microbiome data could be used to predict the culture-based categorisation of nasal \u003cem\u003eS. aureus\u003c/em\u003e colonisation status. Additionally, this served as a sensitivity analysis for the previous differential abundance analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF), which allows for the identification of significant microbial determinants for \u003cem\u003eS. aureus\u003c/em\u003e colonisation status. We split the data into training and test data at a ratio of 80:20, and determined the best number of candidates to be sampled at each tree (mtry) to be 6. The estimated test classification accuracy of the trained model was 73.2% (1-estimated out of box error) with the lowest class error for non-carriers (6.85%) and highest for intermittent carriers (100%).\u003c/p\u003e \u003cp\u003eWe determined the accuracy, sensitivity and specificity of the model with the test data. The overall accuracy of the model was 75.2% (C.I.= 67.4%-81.9%, p\u0026thinsp;=\u0026thinsp;\u0026lt;\u0026thinsp;0.001) significantly exceeding the no information rate (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). Overall, the model performed best in predicting persistent colonisation with 83.0% and 88.2% sensitivity and specificity, respectively, suggesting the greatest utility for identification of individuals at higher risk of persistent \u003cem\u003eS. aureus\u003c/em\u003e colonisation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). For non-carriers, the sensitivity was higher at 94.8%, but specificity lower at 66.6%. For intermittent carriers the sensitivity was 0.0% suggesting the model was completely unable to predict the intermittent colonisation from the microbiome data; of the 25 intermittent carriers in the test dataset, none were classified as intermittent carriers, 16/25 (64%) were misclassified as non-carriers and 9/25 (36%) as persistent carriers, adding further evidence that intermittent carriers are not distinct group, and a greater proportion are similar to non-carriers compared to persistent carriers.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe determined variable importance (i.e. how much each variable contributes to the prediction) by evaluating the mean decrease in accuracy (a measure of decrease in the model accuracy computed by permuting out-of-box error data) and the mean decrease in gini index (a measure of variance and resulting misclassification across the random forest nodes) after removal of each feature, i.e. taxon \u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. The top three features of importance by assessing the mean decrease in accuracy were \u003cem\u003eS. aureus\u003c/em\u003e, \u003cem\u003eCorynebacterium sp.\u003c/em\u003e, and \u003cem\u003eS. epidermidis\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). The top three features of importance by assessing the mean decrease in gini index were \u003cem\u003eS. aureus\u003c/em\u003e, \u003cem\u003eCorynebacterium sp.\u003c/em\u003e, and \u003cem\u003eD. pigrum\u003c/em\u003e, with \u003cem\u003eS. aureus\u003c/em\u003e clearly contributing the most to the model (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003e \u003cb\u003eStaphylococcus aureus\u003c/b\u003e \u003cb\u003ephylogenetic associations with carriage\u003c/b\u003e\u003c/p\u003e \u003cp\u003eWe next investigated if certain \u003cem\u003eS. aureus\u003c/em\u003e lineages have a propensity for persistent nasal carriage or are more capable of dominance of the community compared to other competing resident bacteria. We used \u003cem\u003eS. aureus\u003c/em\u003e isolate whole genome sequences with matched microbiome data (n\u0026thinsp;=\u0026thinsp;172) and compared the \u003cem\u003eS. aureus\u003c/em\u003e phylogenetic tree, and major multi-locus sequence types (MLST), to the colonisation state and the sample microbiome (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). Following the bifurcation at the root of the phylogeny, as seen in large collections of diverse \u003cem\u003eS. aureus\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, there is a greater number of samples showing higher \u003cem\u003eS. aureus\u003c/em\u003e abundance amongst isolates in cluster B (dominated by ST30, ST34, ST398, and ST45) with a lower number of samples showing higher abundance of species identified earlier as showing a negative association with \u003cem\u003eS. aureus\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF). We examined differences in rarefied \u003cem\u003eS. aureus\u003c/em\u003e abundance (n\u0026thinsp;=\u0026thinsp;125), which demonstrated a significantly higher abundance in samples in cluster B compared to cluster A (Mann-Whitney, p\u0026thinsp;=\u0026thinsp;0.041) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). For the \u003cem\u003eS. aureus\u003c/em\u003e dominant CST1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), 38/74 (51.4%) of cluster A compared to 33/51 (64.7%) cluster B samples were found in this CST. Next, we assessed differences in Beta diversity between cluster A and B (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC, D), and found a small but statistically significant (PERMANOVA analysis (\u003cem\u003eF\u003c/em\u003e(2)\u0026thinsp;=\u0026thinsp;2.33, p\u0026thinsp;=\u0026thinsp;0.04) divergence of these groups.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eDespite the importance of \u003cem\u003eS. aureus\u003c/em\u003e colonisation as a risk factor for \u003cem\u003eS. aureus\u003c/em\u003e infection, there is still only a limited understanding of what determines nasal \u003cem\u003eS. aureus\u003c/em\u003e colonisation. In this work, we combine for the first time, large-scale microbiome sequencing with longitudinal culture data that, since the 1940\u0026rsquo;s \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, has been used to define \u003cem\u003eS. aureus\u003c/em\u003e colonisation. We have generated multiple new insights into the nasal microbial community structure of the anterior nares, substantially extending previous smaller-scale studies \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. Like the previous study of older twins from Denmark \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, we identify seven community state types (CSTs), but with a different species composition which suggests that either the previous smaller study was unrepresentative or there is variation in the nasal microbiota even between two northern European countries. Importantly, our analysis of the seven CSTs revealed that women are more likely to have either CST 6 (\u003cem\u003eC. accolens\u003c/em\u003e) or CST 7 (diverse group), suggesting an influence of sex on the wider nasal microbiome composition, as with \u003cem\u003eS. aureus\u003c/em\u003e persistent colonisation \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e and load \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe demonstrate that there is a clear distinction in the microbial community structure that underlies \u0026lsquo;persistent\u0026rsquo; \u003cem\u003eS. aureus\u003c/em\u003e carriage compared to non-\u003cem\u003eS. aureus\u003c/em\u003e carriage in a large sample of individuals (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eE). We found that persistent carriage of \u003cem\u003eS. aureus\u003c/em\u003e is negatively associated with three \u003cem\u003eCorynebacterium\u003c/em\u003e species (including \u003cem\u003eC. jeikeiu\u003c/em\u003em, \u003cem\u003eC. accolens\u003c/em\u003e and an unnamed \u003cem\u003eCorynebacterium sp\u003c/em\u003e), \u003cem\u003eD. pigrum\u003c/em\u003e, \u003cem\u003eS. epidermidis\u003c/em\u003e, and \u003cem\u003eM. catarrhalis\u003c/em\u003e. Notably, \u003cem\u003eC. jeikeium\u003c/em\u003e, \u003cem\u003eC. accolens, M. catarrhalis\u003c/em\u003e, and the unnamed \u003cem\u003eCorynebacterium sp.\u003c/em\u003e have not been previously identified as negatively associated with \u003cem\u003eS. aureus\u003c/em\u003e abundance in microbiome data \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e,\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. We failed to replicate the negative association with \u003cem\u003eS. aureus\u003c/em\u003e abundance with \u003cem\u003eSimonsiella sp.\u003c/em\u003e or \u003cem\u003eCutibacterium\u003c/em\u003e (formerly \u003cem\u003ePropionibacterium) granulosum\u003c/em\u003e as previously reported \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. Neither species were found in our pre or post-QC data, suggesting that these are either uncommon species in England, perhaps only present in certain environmental conditions, or contaminants \u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn contrast, our findings did replicate the negative association previously reported for \u003cem\u003eDolosigranulum\u003c/em\u003e spp. \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e and \u003cem\u003eD. pigrum\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Interactions between \u003cem\u003eD. pigrum and Corynebacterium\u003c/em\u003e spp. and particularly \u003cem\u003eC. accolens\u003c/em\u003e have recently been explored \u003cem\u003ein vitro\u003c/em\u003e; \u003cem\u003eD pigrum\u003c/em\u003e was shown to inhibit a single strain of \u003cem\u003eS. aureus\u003c/em\u003e directly, whilst \u003cem\u003eC. pseudodipthericum\u003c/em\u003e, \u003cem\u003eC. accolens\u003c/em\u003e, \u003cem\u003eC. propinquum\u003c/em\u003e conditioned media enhanced the growth of \u003cem\u003eD. pigrum\u003c/em\u003e, and that this growth enhancement was not reciprocal \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. \u003cem\u003eC. accolens\u003c/em\u003e was found to both enhance the growth of \u003cem\u003eD. pigrum\u003c/em\u003e through an unknown mechanism(s), and inhibit growth by processing host tri-acylglycerols into fatty acids with antibacterial properties \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. Our data further highlights the extent of the mutualism between taxa that are negatively associated with \u003cem\u003eS. aureus\u003c/em\u003e carriage such as \u003cem\u003eD. pigrum\u003c/em\u003e and multiple \u003cem\u003eCorynebacterium\u003c/em\u003e spp. (see co-occurrence in CST2, 3 and 4 in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea), suggesting that this is an area which needs further systematic investigation.\u003c/p\u003e \u003cp\u003eWe found both \u003cem\u003eC. jeikeium and C. accolens\u003c/em\u003e are negatively associated with \u003cem\u003eS. aureus\u003c/em\u003e. While \u003cem\u003eC. jeikeium\u003c/em\u003e has not been previously reported to inhibit \u003cem\u003eS. aureus\u003c/em\u003e, \u003cem\u003eC. accolens\u003c/em\u003e has been demonstrated to inhibit \u003cem\u003eS. aureus\u003c/em\u003e experimentally \u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e, and as noted above, \u003cem\u003eC. accolens\u003c/em\u003e can support or inhibit growth of \u003cem\u003eD. pigrum\u003c/em\u003e - which can inhibit \u003cem\u003eS. aureus\u003c/em\u003e directly \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. This suggests \u003cem\u003eC. accolens\u003c/em\u003e can inhibit \u003cem\u003eS. aureus\u003c/em\u003e through direct or indirect means. This is supported by our data whereby most individuals with \u003cem\u003eC. accolens\u003c/em\u003e are found in CST6, which is dominated by \u003cem\u003eC. accolens\u003c/em\u003e and has a low abundance of \u003cem\u003eS. aureus\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Notably, women have an increased relative risk for CST6; this provides a mechanistic explanation for why women are less likely to be persistent carriers of \u003cem\u003eS. aureus\u003c/em\u003e, although this is unlikely to explain the entire variation in colonisation rates between men and women \u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e. As discussed previously, \u003cem\u003eC. accolens\u003c/em\u003e has also been reported to be positively associated with \u003cem\u003eS. aureus\u003c/em\u003e in a smaller microbiome study and was shown to promote the growth of \u003cem\u003eS. aureus in vitro\u003c/em\u003e \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. In our data, \u003cem\u003eC. accolens\u003c/em\u003e is found at a lower abundance in multiple CSTs, including co-occurrence with \u003cem\u003eS. aureus\u003c/em\u003e in a small sub-cluster within the \u003cem\u003eS. aureus\u003c/em\u003e dominated CST1 (see CST1 in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). This suggests that there is a strain level or lineage variation in \u003cem\u003eC. accolens\u003c/em\u003e and/or \u003cem\u003eS. aureus\u003c/em\u003e that contributes to this discordant relationship \u0026ndash; which again highlights the need for large scale studies.\u003c/p\u003e \u003cp\u003eOur data identified a negative association between \u003cem\u003eS. epidermidis\u003c/em\u003e and \u003cem\u003eS. aureus\u003c/em\u003e, corroborated by the differential abundance analysis and the random forest model. Although negative \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e and positive \u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e associations have been reported previously, similar to \u003cem\u003eC. accolens\u003c/em\u003e, a previous study has shown that a certain proportion of \u003cem\u003eS. epidermidis\u003c/em\u003e strains secrete a serine protease, Esp, which inhibits \u003cem\u003eS. aureus\u003c/em\u003e growth \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Furthermore, the effect observed could also be explained by environmental factors. \u003cem\u003eS. aureus\u003c/em\u003e requires a higher relative humidity (87%) than \u003cem\u003eS. epidermidis\u003c/em\u003e (81%), and \u003cem\u003eS. epidermidis\u003c/em\u003e may be more commonly found in \"drier\" noses \u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Our finding that \u003cem\u003eM. catarrhalis\u003c/em\u003e is negatively associated with \u003cem\u003eS. aureus\u003c/em\u003e in adults is new, though abundance has been reported to be inversely correlated with \u003cem\u003eS. aureus\u003c/em\u003e in children \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAs would be expected, persistent carriage was positively associated with \u003cem\u003eS. aureus\u003c/em\u003e. Strikingly, we found that in ~\u0026thinsp;50% of persistently colonised individuals \u003cem\u003eS. aureus\u003c/em\u003e is the single most abundant organism in the nasal microbiome, representing\u0026thinsp;\u0026gt;\u0026thinsp;75% of reads for ~\u0026thinsp;35% persistent carriers, indicating that when present it dominates the niche. This domination is reflected in reduced Alpha diversity among persistent carriers compared to non-carriers and is further supported by the greater stability of Alpha diversity amongst the persistent \u003cem\u003eS. aureus\u003c/em\u003e carriers compared to non-carriers. This suggests that \u003cem\u003eS. aureus\u003c/em\u003e may act as a keystone species which principally determines its own continued carriage and can suppress other members of the nasal microbiome. While most nasal microbiome research has concentrated on how other members of the nasal microbiome prevent \u003cem\u003eS. aureus\u003c/em\u003e colonisation, our data indicates that future investigations should now focus how \u003cem\u003eS. aureus\u003c/em\u003e excludes other species from the nasal microbiome and is resistant to antagonistic compounds or growth conditions generated by competing species such as \u003cem\u003eCorynebacterium\u003c/em\u003e spp. and \u003cem\u003eD. pigrum\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eThe subset of very high \u003cem\u003eS. aureus\u003c/em\u003e load carriers we identified might also be of clinical importance: The high bacterial load of \u003cem\u003eS. aureus\u003c/em\u003e and a lack of antagonistic species amongst persistent carriers specifically may explain why these individuals are more likely to become infected by \u003cem\u003eS. aureus\u003c/em\u003e; inoculation of a wound with a high load of \u003cem\u003eS. aureus\u003c/em\u003e in the absence antagonistic species (that might inhibit \u003cem\u003eS. aureus\u003c/em\u003e) could make infection is more likely. Future studies which make the use of nasal microbiome-based stratified participants are required to understand the consequences of high-load \u003cem\u003eS. aureus\u003c/em\u003e dominant nasal colonisation on the rest of the human microbiome (e.g. skin and gut \u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e) and subsequent risk of infection.\u003c/p\u003e \u003cp\u003eImportantly, our study demonstrates that \u003cem\u003eS. aureus\u003c/em\u003e carriage can be predicted from microbiome data with a moderate degree of accuracy. Notably, the model is more sensitive in predicting \u003cem\u003eS. aureus\u003c/em\u003e non-carriage. This is particularly important given a single swab of the anterior nares is limited in its diagnostic accuracy \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Further, the high negative predictive value of the model for persistent carriage we present here, using the microbiome data from a single swab, may improve the identification of true negatives in \u003cem\u003eS. aureus\u003c/em\u003e screening and identify those who are unlikely to be persistently colonised, facilitating a selective approach to patient decolonisation. Further large-scale studies using higher resolution metagenomics and clinical data will likely significantly improve this, moving towards being able to define risk of \u003cem\u003eS. aureus\u003c/em\u003e infection based on a single swab.\u003c/p\u003e \u003cp\u003eWe have determined that \u0026lsquo;intermittent\u0026rsquo; carriers do not have a distinct microbial community. This was first proposed by Van Belkum \u003cem\u003eet al\u003c/em\u003e. \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e who theorised that: (1) non-carriage is either incidental and most people are actually intermittent carriers or (2) intermittent carriers are non-carriers who have picked up \u003cem\u003eS. aureus\u003c/em\u003e from the environment. Our data provides evidence that the latter of the two is partly correct and provides new insights into this. Intermittent carriers, as defined by one or two culture positive swabs, belong to one of two populations: a population with a \u003cem\u003eS. aureus\u003c/em\u003e dominated microbial community structure (similar to persistent carriers) with the absence of species identified as negatively associated with \u003cem\u003eS. aureus\u003c/em\u003e or a population with one of several microbial community structures which are not dominated by \u003cem\u003eS. aureus\u003c/em\u003e, are more diverse, and often dominated by other species (similar to non-carriers) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee). Therefore, the most parsimonious explanation is that \u0026lsquo;intermittent\u0026rsquo; carriers, given their low \u003cem\u003eS. aureus\u003c/em\u003e abundance and predominance amongst non-carriage CSTs, are effectively \u003cem\u003eS. aureus\u003c/em\u003e \u0026lsquo;non-dominant carriers\u0026rsquo; who are only transiently colonised with \u003cem\u003eS. aureus\u003c/em\u003e, for example due to environmental/household exposure (i.e. hypothesis 2 proposed by Van Belkum \u003cem\u003eet al\u003c/em\u003e.). While two swab positive \u0026lsquo;intermittent\u0026rsquo; carriers are \u0026lsquo;\u003cem\u003eS. aureus\u003c/em\u003e dominant carriers\u0026rsquo; (akin to persistent carriers), reflected by their higher \u003cem\u003eS. aureus\u003c/em\u003e abundance and greater representation amongst the \u003cem\u003eS. aureus\u003c/em\u003e dominant CST, and therefore likely to be persistently colonised individuals that were negative by culture in one swab.\u003c/p\u003e \u003cp\u003eWe identified a relationship between certain \u003cem\u003eS. aureus\u003c/em\u003e lineages and \u003cem\u003eS. aureus\u003c/em\u003e abundance and the associated microbiome. This suggests that \u003cem\u003eS. aureus\u003c/em\u003e abundance and carriage is to some degree lineage specific. Previous studies of human experimental colonisation with \u003cem\u003eS. aureus\u003c/em\u003e identified that after decolonisation and artificial inoculation, persistent carriers had higher loads of \u003cem\u003eS. aureus\u003c/em\u003e than intermittent or non-carriers and were more likely to select their own strain \u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Our data suggest that this might be due to lineage-specific effects of persistent carriers\u0026rsquo; strains being better adapted to persistent colonisation, enabling them to reach higher abundances. The same study also showed that persistent carriers had higher serum IgG and IgA levels to certain \u003cem\u003eS. aureus\u003c/em\u003e antigens (SasG, TSST-1, SEA, ClfA, CHIPS). Given the known variation of mobile genetic element (MGE) content in \u003cem\u003eS. aureus\u003c/em\u003e lineages (TSST-1, SEA, CHIPS are all MGE acquired), a propensity for lineages with particular MGE content to be found at higher abundance in persistent carriers may explain this variation in antibody levels. Overall, these lineage relationships are likely to represent adaptations of \u003cem\u003eS. aureus\u003c/em\u003e \u003csup\u003e\u003cem\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/em\u003e\u003c/sup\u003e that impact host adaption to colonisation, transmission success and intra- and interspecies competition. This requires further investigation to further elucidate the mechanisms for this lineage specific adaption to colonisation.\u003c/p\u003e \u003cp\u003eThis study has potential limitations, including those inherent to 16S rRNA gene studies; though we have previously demonstrated the accuracy of our sequencing and analysis pipeline \u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e, further validation of taxa (for example: the unnamed \u003cem\u003eCorynebacterium sp.\u003c/em\u003e negatively associated with \u003cem\u003eS. aureus\u003c/em\u003e) identified at a species level with selective culture and/or shotgun metagenomic sequencing will be useful. Additionally, although the study participants are healthy and sampled at home, the blood donor cohorts used will still not be entirely representative of the population of England, which is itself clearly not representative of all global populations. Further studies in different populations using standardised methods are now required to explore this variation.\u003c/p\u003e \u003cp\u003eIn summary, we present the most comprehensive assessment of the microbial composition of the anterior nares to date. Our data provides multiple new insights and identifies key microbial interactions and variation that underpin the composition of the human nasal microbiome, and in particular colonisation by \u003cem\u003eS. aureus.\u003c/em\u003e\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe CARRIAGE study protocol was approved by the National Research Ethics Service Committee North-West - Lancaster Research Ethics Committee, 27/06/2016, REC reference: 16/NW/0507, IRAS project ID: 202688. All participants provided informed consent. The study is registered at ISRCTN: ISRCTN10474633\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eParticipants and samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this observational cohort study, nasal samples taken from the anterior nares were obtained from healthy human participants from the community participating in the CARRIAGE study \u003csup\u003e49\u003c/sup\u003e between 13\u003csup\u003eth\u003c/sup\u003e October 2016 and 17\u003csup\u003eth\u003c/sup\u003e May 2017. Briefly, \u003cem\u003eS. aureus\u003c/em\u003e colonisation status was assessed by culture of three self-administered nasal swabs delivered to participants and taken at weekly intervals, and subsequently posted back to the laboratory. \u003cem\u003eS. aureus\u003c/em\u003e colonisation status was defined as: (i) persistent colonisation, based on three \u003cem\u003eS. aureus\u003c/em\u003e culture positive weekly nasal swabs, (ii) intermittent colonisation, defined as one or two swabs positive, and (iii) non-carrier status, defined as no swabs positive, based on previous studies \u003csup\u003e9,10,33,34\u003c/sup\u003e (\u003cstrong\u003eFig. 1A\u003c/strong\u003e). \u003c/p\u003e\n\u003cp\u003eLifestyle information was collected by questionnaires or from pre-existing data held as part of baseline questionnaires in previous studies involving the same participants. The Amies transport liquid that the swabs (the same swabs that were used for culture) were transported to the laboratory in were processed without culture for 16S rRNA gene sequencing to identify the microbial community composition (\u003cstrong\u003eSupplementary Fig. S1 and 2\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eS. aureus\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e culture\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAfter 10 seconds of vortexing, nasal swabs in Amies transport media (Medical Wire) were transferred to a tube containing 2\u0026thinsp;ml Tryptic Soy broth supplemented with 6.5 % NaCl (Medical Wire) and incubated overnight at 37\u0026thinsp;\u0026deg;C, in air. The remaining Amies solution was transferred to an Eppendorf with 500 \u0026micro;l of glycerol, pipette mixed and stored at -70\u0026ordm;C. 10\u0026thinsp;\u0026micro;l of the overnight enrichment broth was streaked onto chromogenic Staph Brilliance 24 agar plates (Oxoid) and incubated overnight at 37\u0026thinsp;\u0026deg;C. If no blue colonies were identified after 24 hours of incubation, the plate was returned to the incubator overnight and rechecked. Blue colonies are with the phenotypes of putative \u003cem\u003eS. aureus\u003c/em\u003e were sub-cultured onto Columbia Blood agar (5% horse blood) and incubated overnight at 37\u0026thinsp;\u0026deg;C. Colonies from these plates were inspected visually for phenotype indicators, and tested for coagulase and protein A via latex agglutination test (Pro-Lab Diagnostic). Where there were queries or discrepancies, species level identity was confirmed using Matrix assisted laser desorption and ionisation \u0026ndash; Time of Flight (MALDI-ToF). All isolates were stored (Pro-Lab Diagnostics) at -80\u0026thinsp;\u0026deg;C.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDNA processing and 16S rRNA gene polymerase chain reaction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePrior to extraction, residual sample transport medium from nasal samples was stored at -70\u0026ordm;C in ~33% v/v glycerol. Total DNA was extracted from nasal sample transport medium after an additional mechanical lysis step (Lysing matrix E, MP Biomedicals) either via the MPBio MPure-12 instrument, (MPure Bacterial DNA Kit, MP Biomedicals) or manually using the FastSpin Kit for Soil (MPBiomedicals), including the heated elution step. DNA was then stored at -70\u0026ordm;C until use. V1V2 specific primers with attached sequencing adaptors and indexes (\u003cstrong\u003eTable S1\u003c/strong\u003e) were used for PCR to amplify the bacterial 16S ribosomal gene regions. PCR amplification mastermixes were prepared manually using a Q5 High-Fidelity Polymerase Kit (M0491, New England Biolabs). PCR amplifications were setup in triplicate (25ul each), products were pooled into a single volume per sample, and all samples were subsequently purified using an AMPure XP (Beckman Coulter) workflow at a ratio of 0.8X. Libraries were quantified using the Qubit HS DNA Kit (ThermoFisher). Equimolar pools were then created. Negative controls included a sample extraction control, a PCR water control, and an aliquot of the glycerol used for storage, whilst a positive control was represented by purified water spiked with \u003cem\u003eS. aureus \u003c/em\u003eDNA. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDNA sequencing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePer experiment, an equimolar pool of PCR libraries was sequenced at the Wellcome Sanger Institute in-house sequencing facility, using the Illumina MiSeq (300bp paired-end reads, v3 Reagent Kit). Accession numbers for the sequencing data is in \u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Table S6\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e16S rRNA gene sequence quality control and taxonomy assignment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe used a modified mothur MiSeq standard operating procedure (SOP) to process paired fastq files (MOTHUR wiki at http://www.mothur.org/wiki/MiSeq_SOP) \u003csup\u003e50\u003c/sup\u003e. The four poly(NNNN)s present in the adapter/primer sequences of contigs assembled with the \u0026lsquo;make.contigs\u0026rsquo; command in mothur were trimmed with the PRINSEQ program, before the modified MiSeq SOP was resumed. The Silva bacterial database \u0026lsquo;silva.nr_v132.align\u0026rsquo; was used to align quality-screened sequences and chimeras removed using Uchime\u003csup\u003e51\u003c/sup\u003e. Sequences were then classified using the same Silva reference database and the Silva taxonomy database \u0026lsquo;silva.nr_v132.tax\u0026rsquo;, with the removal of chloroplast, mitochondria, unknown, and eukaryota sequences. We clustered high-quality unique sequences with Oligotyping v2.1\u003csup\u003e52\u003c/sup\u003e (-M option to 1000), which were assigned to NODES, and referred to as operational taxonomic units (OTU) from here, with the \u0026lsquo;Minimum Entropy Decomposition\u0026rsquo; (MED) option. We created a customised silva SSU Ref database (NR99, release 132), where we removed the majority of environmental and uncultured taxa, and carried out taxonomic assignment with ARB (v6.0.6-3)\u003csup\u003e53\u003c/sup\u003e. In some instances, where a mismatch was observed within the taxonomic groups, we assigned taxa to the OTU sequence with BLAST \u003csup\u003e54\u003c/sup\u003e (see \u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Table S2\u003c/strong\u003e). We then combined the output in R (v4.4.1) into a phyloseq \u003csup\u003e55\u003c/sup\u003e object for onward analysis. \u003c/p\u003e\n\u003cp id=\"_Toc163598272\"\u003e\u003cstrong\u003eContaminant removal and accounting for variability in sequencing depth\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe identified contaminants and removed these by identifying batch effects and accounting for negative controls \u003csup\u003e28-30\u003c/sup\u003e. Batch effects were assessed by calculating the spearman\u0026rsquo;s correlation co-efficient of species against each location of extraction, and location of PCR reaction. We then examined correlation of species with sample DNA concentrations. We used well characterised \u0026lsquo;kitome\u0026rsquo; and environment contaminants to identify additional associated contaminants by calculating \u0026lsquo;species-species\u0026rsquo; correlation coefficients. We used Decontam v1.16.0\u003csup\u003e56\u003c/sup\u003e to account for laboratory negative controls, run with the \u0026lsquo;isnotcontam\u0026rsquo; function and with each sequencing run provided as a batch (further details below and in \u003cstrong\u003eSupplementary Fig. S3 and S4 and Supplementary Table S3\u003c/strong\u003e)\u003c/p\u003e\n\u003cp\u003eWe determined a suitable rarefication depth of 10,000 reads using rarefication curves and examining the read depth at which the majority of sample taxa numbers plateaued (see rarefaction section below and \u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig. S5\u003c/strong\u003e). We removed species with an abundance of less than 0.1% across samples, below which we expected the removal of most contaminants and account for the variability in rare species composition between runs \u003csup\u003e48\u003c/sup\u003e. For diversity analyses, the rarified dataset was used. For abundance analyses, to mitigate data loss, we combined samples with greater than 500 high quality reads with samples that had greater than 10000 reads and rarefied. (see rarefaction section, \u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig. S3, S4, S5 and \u003c/strong\u003e\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Table S3\u003c/strong\u003e). \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIdentification of contaminants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRemoval of taxa below the 0.1% threshold resulted in 115/2,322 OTUs remaining. The samples were processed over two time periods. Over the first time period (n=1,099), there were five locations for DNA extractions and five locations for PCR amplification. we used spearman\u0026rsquo;s correlation coefficient to identify batch effects, specifically species with abundance that was associated with the extraction and PCR locations (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Table S3, \u003c/strong\u003e\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig., S3 and S4\u003c/strong\u003e). For the second time period (n=767), extractions and PCR amplifications took place in one location and therefore batch effects by location was not examined. We used spearman\u0026rsquo;s correlation coefficient to identify taxa that correlated with PCR qubit values (post-PCR amplification DNA concentration); previously, lower sample DNA concentrations have been associated with contaminants \u003csup\u003e30,57\u003c/sup\u003e (\u003cstrong\u003eTable S3, \u003c/strong\u003e\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig., S3 and S4\u003c/strong\u003e). We used hierarchical clustering to identify species that clustered with one another, which allowed for the identification of taxa that were correlated with well-characterised and suspected contaminants \u003csup\u003e30\u003c/sup\u003e (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Table S3, \u003c/strong\u003e\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig., S3 and S4\u003c/strong\u003e). As a final check, we used the R package Decontam (v1.16.0) \u003csup\u003e56\u003c/sup\u003e to account for negative controls, with each sequencing run considered as a batch (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Table S3, Fig. S3 and S4\u003c/strong\u003e). \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDetermining a rarefication threshold\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe subset CARRIAGE samples and generated rarefication curves for samples with greater than 1,000, 5,000, 10,000, 15,000, 20,000, 100,000 high-quality reads respectively (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig. S5\u003c/strong\u003e). In order to determine a rarefication threshold, we identified the slope of each rarefication curve at the respective high-quality read threshold using the rareslope() function in phyloseq (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig. S5\u003c/strong\u003e); given the large dataset, visualising the point at which the curves plateaued was not possible. It was apparent that at greater high-quality read thresholds, a larger proportion of the samples reached a (near) plateau. We aimed to reach a balance between losing a large number of samples and retaining a dataset where the rarefication curves for the vast majority of samples had plateaued; this was met at 10,000 reads (\u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Fig. S5\u003c/strong\u003e). From here, we either use the dataset rarefied to an even-depth to the minimum read count above this threshold (10,004) or this dataset combined with samples with greater than 500 reads but less than 10,000 reads, to minimise data loss and consistent with previous analyses\u003csup\u003e58\u003c/sup\u003e. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDiversity analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted microbial diversity and compositional analysis in R using diversity indices calculated with the phyloseq (v1.40) \u003csup\u003e55\u003c/sup\u003e and vegan (v2.6-4) \u003csup\u003e59\u003c/sup\u003e packages. Alpha-diversity indices (Shannon\u0026rsquo;s and Simpson\u0026rsquo;s) were calculated on rarefied read counts. Sample microbial composition is consistently represented with relative abundances. We used Principal Coordinate Analysis (PCoA) and Non-Metric Dimensional Scaling (NMDS) with the bray\u0026ndash;curtis distance matrices to visualise differences in sample diversity by condition (e.g. \u003cem\u003eS. aureus\u003c/em\u003e colonisation status). \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData visualisation and statistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe manipulated data in Excel 2016 and R (v4.4.1). We generated figures using ggplot2 (v3.4.0), phyloseq (v1.40) \u003csup\u003e55\u003c/sup\u003e, and microViz (v0.11.0). We evaluated differences in Alpha indices with Mann-Whitney-U and Kruskall-Wallis tests where appropriate. We used PERMANOVA to estimate differences between Bray-Curtis distances observed by study groups with the vegan package (v2.6-4)\u003csup\u003e59\u003c/sup\u003e. To determine the number of clusters in the data, we calculated a gap statistic with ordination values using Bray-Curtis distances, using the R package \u0026lsquo;cluster\u0026rsquo; function clusGap() (\u003cstrong\u003eFig. S6\u003c/strong\u003e). We investigated the association of plausible lifestyle and comorbidities risk factors with Community State Types (CST) using a multinomial logistic regression model analysis (CST ~ sex + smoking status + pet ownership + healthcare contact + chronic skin condition + asthma + allergies). We used ANCOM-BC (v1.6.4) \u003csup\u003e60\u003c/sup\u003e to evaluate differential abundance of microbial species in the study groups; we used the ancombc2 function with default settings, but specified taxa with a prevalence of less than 0.1% to be removed and a library cut-off of 500 reads, and provided a non-rarefied count table as a centred log ratio transformation is conducted \u003csup\u003e60\u003c/sup\u003e. This study complies with the STORMS guidelines \u003csup\u003e61\u003c/sup\u003e for reporting.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRandom forest model\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe trained two separate models, one utilising all samples above 500 reads where samples with greater than 10,004 had been rarefied (n=1,055), and another including the rarefied dataset alone (n=795). The rarefied dataset performed better compared to the combined dataset (see Supplementary Results for further details). We used the R package randomForest (v4.7-1.1)\u003csup\u003e37\u003c/sup\u003e to fit a random forest classifier for carriage status (relative_microbial_abundance_data ~ carriage_status). The model was trained using a randomly subsampled dataset of the microbial features (in relative abundance format) representing 80% of the data (ntrees=1000), and tested on the remaining 20% to evaluate model robustness. We chose the number of predictors sampled for splitting at each node (mtry) with the tuneRF() function. We obtained sensitivity and specificity values of the model with the R package caret (v6.0-90) whilst receiver operating characteristic curve (ROC) curves and AUC were obtained with the R package pROC (v1.18.4). P-values less than 0.05 were considered statistically significant. \u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWhole Genome Sequencing of \u003cem\u003eStaphylococcus aureus\u003c/em\u003e isolates\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eS. aureus\u003c/em\u003e isolates were sequenced at the Wellcome Sanger Institute with 96 sample libraries sequenced on a 300bp PE MiSeq lane (with a 1% PhiX spike). European Nucleotide Accession number for isolates is presented in \u003cstrong\u003eSupplementary\u003c/strong\u003e\u003cstrong\u003e Table S7\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePhylogenetic analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFrom the raw whole genome sequencing data, we generated quality control metrics, and trimmed reads, with the nextflow pipelines, bacQC (github.com/avantonder/bacQC). Species classification for each sample was performed using Kraken and Bracken\u003csup\u003e62\u003c/sup\u003e. We discarded samples with less than 90 % reads matching to \u003cem\u003eS. aureus\u003c/em\u003e and those with \u0026lt;30x coverage from onward analyses. Using the nextflow pipeline, assembleBAC (github.com/avantonder/assembleBAC), we produced annotated assemblies with trimmed fastqs. The pipeline uses shovill (v1.1.0) for assembly. We annotated assemblies with prokka (v.1.14.5) \u003csup\u003e63\u003c/sup\u003e using a genus-specific database from RefSeq for annotation. Assemblies with an N50 value \u0026lt;10000, length of less than 2.6Mbp or greater than 3.0Mbp, or with a spuriously high number of contigs summarised by MultiQC \u003csup\u003e64\u003c/sup\u003e and the QC metrics generated by Panaroo(v1.3.4) \u003csup\u003e65\u003c/sup\u003e were removed from onward analyses. Samples with greater than 300 contigs were found to be outliers. \u003c/p\u003e\n\u003cp\u003eWe assigned sequence types (STs) with mlst (v2.19.0) (github.com/tseemann/mlst), and where these were not assigned, assemblies we queried the sequences on the PubMLST web server \u003csup\u003e66\u003c/sup\u003e. We produced core-genome alignments with Panaroo (v 1.3.4)\u003csup\u003e65\u003c/sup\u003e with a core-genome threshold set to 98 %. We extracted variant sites from the core-genome alignment with snp-sites (v2.5.1) \u003csup\u003e67\u003c/sup\u003e and coupled with associated values for invariant sites to build a maximum likelihood (ML) phylogenetic tree. We used IQ-TREE (v2.1.2)\u003csup\u003e68\u003c/sup\u003e to estimate ML phylogenetic trees with the optimal phylogenetic trees determined by ModelFinder \u003csup\u003e69\u003c/sup\u003e and branch support statistics generated using the ultrafast bootstrap method \u003csup\u003e70\u003c/sup\u003e. \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll sequencing data is publicly available in the European Nucleotide Archive, with details outlined in Supplementary Table S6 and S7.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that all data cleaning and analysis associated with this article were performed using previously published methods, the applications of which are appropriately cited in the corresponding sections in the Methods. No custom code was developed for the aforementioned purposes. Additional code underlying the figures featured are available from the corresponding authors upon request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe wholehearted thank the CARRIAGE study participants for taking part in the CARRIAGE study and providing the samples that were critical to being able to conduct this research.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by Wellcome Collaborative Award in Science (Grant no. 211864/Z/18/Z) to SJP, JP, JD, JAG, EMH. Isaac Newton trust Grant 17.07(1) to EMH, UKRI Innovation Fellowship: MR/S00291X/1 to EMH, Wellcome Grant reference: 220540/Z/20/A, \u0026apos;Wellcome Sanger Institute Quinquennial Review 2021-2026\u0026apos; \u0026ndash; core funding of Wellcome Sanger Institute, Wellcome Clinical PhD Fellowship: 222903/Z/21/Z to DAg. This research was supported by the NIHR Cambridge Biomedical Research Centre (NIHR203312*). The epidemiological coordinating centre of the CARRIAGE study was additionally supported by awards from the NIHR Blood and Transplant Research (5Unit (BTRU) in Donor Health and Behaviour (NIHR203337), NIHR Cambridge BRC (NIHR203312) (*), and by Health Data Research UK (HDRUK2023.0028), which is funded by the UK Medical Research Council, Engineering and Physical Sciences Research Council, Economic and Social Research Council, Department of Health and Social Care (England), Chief Scientist Office of the Scottish Government Health and Social Care Directorates, Health and Social Care Research and Development Division (Welsh Government), Public Health Agency (Northern Ireland), British Heart Foundation and Wellcome. J.D. holds a British Heart Foundation Personal Chair (CH/12/2/29428). *The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care\u0026apos;.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualisation: SJ, EMH, JP, JD, JAG, and DAg.\u003c/p\u003e\n\u003cp\u003eMethodology: DAg, MCdG, KLB, JW, JP, SJP, EMH\u003c/p\u003e\n\u003cp\u003eData curation: DAg, KLB, BB, SB, RH, CP, MRW, CM, SG, CRdS, LS, JB, SD, EJ, MJ, DAn, SI, AM.\u003c/p\u003e\n\u003cp\u003eInvestigation: KLB, BB, KE, PN, SG, CRdS, LS, LL, CR, XB, JB, DAg.\u003c/p\u003e\n\u003cp\u003eSoftware: DAg, MCdG, AvT.\u003c/p\u003e\n\u003cp\u003eResources: JD, AB, ED, MH, SJP, EMH\u003c/p\u003e\n\u003cp\u003eFormal analysis: DAg.\u003c/p\u003e\n\u003cp\u003eValidation: DAg, MCdG.\u003c/p\u003e\n\u003cp\u003eVisualization: DAg, DYKN.\u003c/p\u003e\n\u003cp\u003eWriting\u0026mdash;original draft preparation: DAg, EMH\u003c/p\u003e\n\u003cp\u003eWriting\u0026mdash;review and editing: All authors.\u003c/p\u003e\n\u003cp\u003eProject administration: KLB, BB, DAg, SB, RH, CP, MRW, CM, CC, SD, EJ, MJ, DAn, SI, AM, SJP, EMH.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSupervision: EMH, SJP, JP, JD, MCdG, JW.\u003c/p\u003e\n\u003cp\u003eFunding acquisition: EMH, SJP, JP, JD, JAG, DAg.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.B. reports institutional grants from AstraZeneca, Bayer, Biogen, BioMarin, Bioverativ, Novartis, Regeneron and Sanofi. J.D. serves on scientific advisory boards for AstraZeneca, Novartis, and UK Biobank, and has received multiple grants from academic, charitable and industry sources outside of the submitted work.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMoss B, Squire JR (1948) Nose and skin carriage of Staphylococcus aureus in patients receiving penicillin. Lancet 1:320\u0026ndash;325. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/s0140-6736(48)92088-1\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/s0140-6736(48)92088-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReagan DR et al (1991) Elimination of coincident Staphylococcus aureus nasal and hand carriage with intranasal application of mupirocin calcium ointment. Ann Intern Med 114:101\u0026ndash;106. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.7326/0003-4819-114-2-101\u003c/span\u003e\u003cspan address=\"https://doi.org:10.7326/0003-4819-114-2-101\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYan M et al (2013) Nasal microenvironments and interspecific interactions influence nasal microbiota complexity and S. aureus carriage. Cell Host Microbe 14:631\u0026ndash;640. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.chom.2013.11.005\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.chom.2013.11.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKluytmans J, van Belkum A, Verbrugh H (1997) Nasal carriage of Staphylococcus aureus: epidemiology, underlying mechanisms, and associated risks. Clin Microbiol Rev 10:505\u0026ndash;520. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1128/CMR.10.3.505\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1128/CMR.10.3.505\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWertheim HF et al (2004) Risk and outcome of nosocomial Staphylococcus aureus bacteraemia in nasal carriers versus non-carriers. Lancet 364:703\u0026ndash;705. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/s0140-6736(04)16897-9\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/s0140-6736(04)16897-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evon Eiff C, Becker K, Machka K, Stammer H, Peters G (2001) Nasal carriage as a source of Staphylococcus aureus bacteremia. Study Group. N Engl J Med 344:11\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/nejm200101043440102\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/nejm200101043440102\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBode LG et al (2010) Preventing surgical-site infections in nasal carriers of Staphylococcus aureus. N Engl J Med 362:9\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1056/NEJMoa0808939\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1056/NEJMoa0808939\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWertheim HF et al (2005) The role of nasal carriage in Staphylococcus aureus infections. Lancet Infect Dis 5:751\u0026ndash;762. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/s1473-3099(05)70295-4\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/s1473-3099(05)70295-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNouwen JL et al (2004) Predicting the Staphylococcus aureus nasal carrier state: derivation and validation of a culture rule. Clin Infect diseases: official publication Infect Dis Soc Am 39:806\u0026ndash;811. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1086/423376\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1086/423376\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Belkum A et al (2009) Reclassification of Staphylococcus aureus nasal carriage types. J Infect Dis 199:1820\u0026ndash;1826. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1086/599119\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1086/599119\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCole AL et al (2018) Cessation from Smoking Improves Innate Host Defense and Clearance of Experimentally Inoculated Nasal Staphylococcus aureus. Infect Immun 86. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1128/IAI.00912-17\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1128/IAI.00912-17\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSollid JU, Furberg AS, Hanssen AM, Johannessen M (2014) Staphylococcus aureus: determinants of human carriage. Infect Genet evolution: J Mol Epidemiol evolutionary Genet Infect Dis 21:531\u0026ndash;541. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.meegid.2013.03.020\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.meegid.2013.03.020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScheuch M et al (2019) Staphylococcus aureus colonization in hemodialysis patients: a prospective 25 months observational study. BMC Nephrol 20:153. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s12882-019-1332-z\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s12882-019-1332-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNguyen MH et al (1999) Nasal carriage of and infection with Staphylococcus aureus in HIV-infected patients. Ann Intern Med 130:221\u0026ndash;225. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.7326/0003-4819-130-3-199902020-00026\u003c/span\u003e\u003cspan address=\"https://doi.org:10.7326/0003-4819-130-3-199902020-00026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMulcahy ME, McLoughlin RM (2016) Host-Bacterial Crosstalk Determines Staphylococcus aureus Nasal Colonization. Trends Microbiol 24:872\u0026ndash;886. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.tim.2016.06.012\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.tim.2016.06.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrugger SD et al (2020) Dolosigranulum pigrum Cooperation and Competition in Human Nasal Microbiota. \u003cem\u003emSphere\u003c/em\u003e 5 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1128/mSphere.00852-20\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1128/mSphere.00852-20\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu CM et al (2015) Staphylococcus aureus and the ecology of the nasal microbiome. Sci Adv 1:e1400216. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1126/sciadv.1400216\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1126/sciadv.1400216\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAccorsi EK et al (2020) Determinants of Staphylococcus aureus carriage in the developing infant nasal microbiome. \u003cem\u003eGenome biology\u003c/em\u003e 21, 301 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://europepmc.org/abstract/MED/33308267 https://genomebiology.biomedcentral.com/track/pdf/10.1186/s13059-020-02209-7\u003c/span\u003e\u003cspan address=\"http://europepmc.org/abstract/MED/33308267 https://genomebiology.biomedcentral.com/track/pdf/10.1186/s13059-020-02209-7\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13059-020-02209-7\u003c/span\u003e\u003cspan address=\"10.1186/s13059-020-02209-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e https://europepmc.org/articles/PMC7731505 https://europepmc.org/articles/PMC7731505?pdf=render\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIwase T et al (2010) Staphylococcus epidermidis Esp inhibits Staphylococcus aureus biofilm formation and nasal colonization. Nature 465:346\u0026ndash;349. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/nature09074\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/nature09074\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZipperer A et al (2016) Human commensals producing a novel antibiotic impair pathogen colonization. Nature 535:511\u0026ndash;516. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/nature18634\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/nature18634\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBomar L, Brugger SD, Lemon KP (2018) Bacterial microbiota of the nasal passages across the span of human life. Curr Opin Microbiol 41:8\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.mib.2017.10.023\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.mib.2017.10.023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Piters S, Binkowska WAA, J., Bogaert D (2020) Early Life Microbiota and Respiratory Tract Infections. Cell Host Microbe 28:223\u0026ndash;232. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.chom.2020.07.004\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.chom.2020.07.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTeo SM et al (2015) The infant nasopharyngeal microbiome impacts severity of lower respiratory infection and risk of asthma development. Cell Host Microbe 17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.chom.2015.03.008\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.chom.2015.03.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBiesbroek G et al (2014) Early respiratory microbiota composition determines bacterial succession patterns and respiratory health in children. Am J Respir Crit Care Med 190:1283\u0026ndash;1292. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1164/rccm.201407-1240OC\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1164/rccm.201407-1240OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUehara Y et al (2000) Bacterial interference among nasal inhabitants: eradication of Staphylococcus aureus from nasal cavities by artificial implantation of Corynebacterium sp. J Hosp Infect 44:127\u0026ndash;133. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1053/jhin.1999.0680\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1053/jhin.1999.0680\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePiewngam P et al (2018) Pathogen elimination by probiotic Bacillus via signalling interference. Nature 562:532\u0026ndash;537. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41586-018-0616-y\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41586-018-0616-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePiewngam P et al (2023) Probiotic for pathogen-specific Staphylococcus aureus decolonisation in Thailand: a phase 2, double-blind, randomised, placebo-controlled trial. Lancet Microbe 4:e75\u0026ndash;e83. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/s2666-5247(22)00322-6\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/s2666-5247(22)00322-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Goffau MC, Charnock-Jones DS, Smith GCS, Parkhill J (2021) Batch effects account for the main findings of an in utero human intestinal bacterial colonization study. Microbiome 9:6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s40168-020-00949-z\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s40168-020-00949-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Goffau MC et al (2019) Human placenta has no microbiome but can contain potential pathogens. Nature 572:329\u0026ndash;334. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41586-019-1451-5\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41586-019-1451-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Goffau MC et al (2018) Recognizing the reagent microbiome. Nat Microbiol 3:851\u0026ndash;853. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41564-018-0202-y\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41564-018-0202-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrank DN et al (2010) The human nasal microbiota and Staphylococcus aureus carriage. PLoS ONE 5:e10598. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1371/journal.pone.0010598\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1371/journal.pone.0010598\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliams RE (1963) Healthy carriage of Staphylococcus aureus: its prevalence and importance. Bacteriological reviews 27:56\u0026ndash;71\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarrison EM et al (2016) Validation of self-administered nasal swabs and postage for the isolation of Staphylococcus aureus. J Med Microbiol 65:1434\u0026ndash;1437. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1099/jmm.0.000381\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1099/jmm.0.000381\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVerhoeven PO et al (2012) An algorithm based on one or two nasal samples is accurate to identify persistent nasal carriers of Staphylococcus aureus. Clin Microbiol infection: official publication Eur Soc Clin Microbiol Infect Dis 18:551\u0026ndash;557. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1111/j.1469-0691.2011.03611.x\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1111/j.1469-0691.2011.03611.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHalablab MA, Hijazi SM, Fawzi MA, Araj GF (2010) Staphylococcus aureus nasal carriage rate and associated risk factors in individuals in the community. Epidemiol Infect 138:702\u0026ndash;706. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1017/S0950268809991233\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1017/S0950268809991233\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndersen PS et al (2013) Risk factors for Staphylococcus aureus nasal colonization in Danish middle-aged and elderly twins. Eur J Clin Microbiol Infect diseases: official publication Eur Soc Clin Microbiol 32:1321\u0026ndash;1326. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1007/s10096-013-1882-0\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1007/s10096-013-1882-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndy Liaw MW \u003cem\u003erandomForest: Breiman and Cutler's random forests for classification and regression\u003c/em\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://rdrr.io/rforge/randomForest/man/importance.html\u003c/span\u003e\u003cspan address=\"https://rdrr.io/rforge/randomForest/man/importance.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichardson EJ et al (2018) Gene exchange drives the ecological success of a multi-host bacterial pathogen. Nat Ecol Evol 2:1468\u0026ndash;1478. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41559-018-0617-0\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41559-018-0617-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCostello EK et al (2009) Bacterial community variation in human body habitats across space and time. Science 326:1694\u0026ndash;1697. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1126/science.1177486\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1126/science.1177486\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIngham AC et al (2021) Dynamics of the Human Nasal Microbiota and Staphylococcus aureus CC398 Carriage in Pig Truck Drivers across One Workweek. Appl Environ Microbiol 87:e0122521. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1128/AEM.01225-21\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1128/AEM.01225-21\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu CM et al (2015) Staphylococcus aureus and the ecology of the nasal microbiome. Sci Adv 1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1126/sciadv.1400216\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1126/sciadv.1400216\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYan M et al (2013) Nasal microenvironments and interspecific interactions influence nasal microbiota complexity and S. aureus carriage. Cell Host Microbe 14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/j.chom.2013.11.005\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/j.chom.2013.11.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTamkin E et al (2025) Airway Corynebacterium interfere with Streptococcus pneumoniae and Staphylococcus aureus infection and express secreted factors selectively targeting each pathogen. Infect Immun 93:e0044524. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1128/iai.00445-24\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1128/iai.00445-24\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang S et al (2022) Corynebacterium accolens inhibits Staphylococcus aureus induced mucosal barrier disruption. Front Microbiol 13:984741. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.3389/fmicb.2022.984741\u003c/span\u003e\u003cspan address=\"https://doi.org:10.3389/fmicb.2022.984741\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Goffau MC, Yang X, van Dijl JM, Harmsen HJ (2009) Bacterial pleomorphism and competition in a relative humidity gradient. Environ Microbiol 11:809\u0026ndash;822. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1111/j.1462-2920.2008.01802.x\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1111/j.1462-2920.2008.01802.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDenis O (2017) Route of transmission of Staphylococcus aureus. Lancet Infect Dis 17:124\u0026ndash;125. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1016/S1473-3099(16)30512-6\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1016/S1473-3099(16)30512-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eColl F et al (2025) The mutational landscape of Staphylococcus aureus during colonisation. Nat Commun 16:302. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41467-024-55186-x\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41467-024-55186-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAggarwal D et al (2023) Optimization of high-throughput 16S rRNA gene amplicon sequencing: an assessment of PCR pooling, mastermix use and contamination. Microb Genom 9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1099/mgen.0.001115\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1099/mgen.0.001115\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u003cem\u003eUnderstanding the biological basis of Staphylococcus aureus CARRIAGE.\u003c/em\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.carriagestudy.org.uk\u003c/span\u003e\u003cspan address=\"https://www.carriagestudy.org.uk\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eP S (2019) \u003cem\u003eMiSeq SOP\u003c/em\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://mothur.org/wiki/miseq_sop/\u003c/span\u003e\u003cspan address=\"https://mothur.org/wiki/miseq_sop/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEdgar RC, Haas BJ, Clemente JC, Quince C, Knight R (2011) UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 27:2194\u0026ndash;2200. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/bioinformatics/btr381\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/bioinformatics/btr381\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEren AM et al (2013) Differentiating between closely related microbial taxa using 16S rRNA gene data. Methods Ecol Evol 4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1111/2041-210X.12114\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1111/2041-210X.12114\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Oligotyping\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLudwig W et al (2004) ARB: a software environment for sequence data. Nucleic Acids Res 32:1363\u0026ndash;1371. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/nar/gkh293\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/nar/gkh293\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorgulis A et al (2008) Database indexing for production MegaBLAST searches. Bioinformatics 24:1757\u0026ndash;1764. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/bioinformatics/btn322\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/bioinformatics/btn322\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcMurdie PJ, Holmes S (2013) phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE 8:e61217. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1371/journal.pone.0061217\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1371/journal.pone.0061217\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavis NM, Proctor DM, Holmes SP, Relman DA, Callahan BJ (2018) Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome 6:226. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s40168-018-0605-2\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s40168-018-0605-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSalter SJ et al (2014) Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol 12:87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s12915-014-0087-z\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s12915-014-0087-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Goffau MC et al (2022) Gut microbiomes from Gambian infants reveal the development of a non-industrialized Prevotella-based trophic network. Nat Microbiol 7:132\u0026ndash;144. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41564-021-01023-6\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41564-021-01023-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOksanen J, Blanchet SG, Kindt F, Legendre R, Minchin P, O'Hara P, Solymos R, Stevens P, Szoecs M, Wagner E, Barbour H, Bedward M, Bolker M, Borcard B, Carvalho D, Chirico G, De Caceres M, Durand M, Evangelista S, FitzJohn H, Friendly R, Furneaux M, Hannigan B, Hill G, Lahti M, McGlinn L, Ouellette D, Ribeiro Cunha M, Smith E, Stier T, Ter Braak A, Weedon C (2022) J. \u003cem\u003e_vegan: Community Ecology Package_. R package version 2.6-4\u003c/em\u003e, \u0026lt; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://CRAN.R-project.org/package=vegan\u0026gt;.\u003c/span\u003e\u003cspan address=\"https://CRAN.R-project.org/package=vegan%3E.\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin H, Peddada SD (2020) Analysis of compositions of microbiomes with bias correction. Nat Commun 11:3514. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41467-020-17041-7\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41467-020-17041-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMirzayi C et al (2021) Reporting guidelines for human microbiome research: the STORMS checklist. Nat Med 27:1885\u0026ndash;1892. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/s41591-021-01552-x\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/s41591-021-01552-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWood DE, Lu J, Langmead B (2019) Improved metagenomic analysis with Kraken 2. Genome Biol 20:257. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s13059-019-1891-0\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s13059-019-1891-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeemann T (2014) Prokka: rapid prokaryotic genome annotation. Bioinformatics 30:2068\u0026ndash;2069. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/bioinformatics/btu153\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/bioinformatics/btu153\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEwels P, Magnusson M, Lundin S, Kaller M (2016) MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32:3047\u0026ndash;3048. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/bioinformatics/btw354\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/bioinformatics/btw354\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTonkin-Hill G et al (2020) Producing polished prokaryotic pangenomes with the Panaroo pipeline. Genome Biol 21:180. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1186/s13059-020-02090-4\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1186/s13059-020-02090-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJolley KA, Bray JE, Maiden MCJ (2018) Open-access bacterial population genomics: BIGSdb software, the PubMLST.org website and their applications. Wellcome Open Res 3:124. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.12688/wellcomeopenres.14826.1\u003c/span\u003e\u003cspan address=\"https://doi.org:10.12688/wellcomeopenres.14826.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePage AJ et al (2016) SNP-sites: rapid efficient extraction of SNPs from multi-FASTA alignments. Microb Genom 2:e000056. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1099/mgen.0.000056\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1099/mgen.0.000056\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinh BQ et al (2020) IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Mol Biol Evol 37:1530\u0026ndash;1534. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/molbev/msaa015\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/molbev/msaa015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKalyaanamoorthy S, Minh BQ, Wong TKF, von Haeseler A, Jermiin LS (2017) ModelFinder: fast model selection for accurate phylogenetic estimates. Nat Methods 14:587\u0026ndash;589. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1038/nmeth.4285\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1038/nmeth.4285\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinh BQ, Nguyen MA, von Haeseler A (2013) Ultrafast approximation for phylogenetic bootstrap. Mol Biol Evol 30:1188\u0026ndash;1195. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org:10.1093/molbev/mst024\u003c/span\u003e\u003cspan address=\"https://doi.org:10.1093/molbev/mst024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6079410/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6079410/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cem\u003eStaphylococcus aureus\u003c/em\u003e colonises the nose in humans, with individuals defined as persistent, intermittent or non-carriers. Unlike the gut microbiome, the nasal microbiome has not been studied in large numbers of people. Here, we define the nasal microbiome in ~ 1,100 individuals and combine this with \u003cem\u003eS. aureus\u003c/em\u003e culture data. We identify seven community state types (CST), including two CSTs found more commonly in females. Approximately 70% of those who are persistently colonised with \u003cem\u003eS. aureus\u003c/em\u003e have a single CST microbiome dominated by \u003cem\u003eS. aureus\u003c/em\u003e, while non-carriers are distributed across the other six CSTs. Intermittent carriers are not a unique state but have microbiomes that resemble either non- or persistent carriers. Persistent carriage is positively associated with \u003cem\u003eS. aureus\u003c/em\u003e abundance, and negatively associated with three \u003cem\u003eCorynebacterium\u003c/em\u003e species, \u003cem\u003eDolosigranulum pigrum\u003c/em\u003e, \u003cem\u003eStaphylococcus epidermidis\u003c/em\u003e, and \u003cem\u003eMoraxella catarrhalis\u003c/em\u003e; the microbiome can be exploited with machine learning to accurately predict \u003cem\u003eS. aureus\u003c/em\u003e colonisation status. Finally, we find that certain \u003cem\u003eS. aureus\u003c/em\u003e lineages are likely better adapted to colonisation. Our data provides a comprehensive view of the nasal microbiome with respect to \u003cem\u003eS. aureus\u003c/em\u003e colonisation, describing two key states: a \u003cem\u003eS. aureus\u003c/em\u003e dominated CST in which \u003cem\u003eS. aureus\u003c/em\u003e shapes the microbiome, and a group of CSTs in which \u003cem\u003eS. aureus\u003c/em\u003e is rare or absent.\u003c/p\u003e","manuscriptTitle":"The nasal microbiome redefines Staphylococcus aureus colonisation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-29 08:24:20","doi":"10.21203/rs.3.rs-6079410/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"9c42f8ea-2877-4594-a8b4-ea85b38f6d9a","owner":[],"postedDate":"April 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47664866,"name":"Biological sciences/Microbiology/Microbial communities/Microbiome"},{"id":47664867,"name":"Biological sciences/Microbiology/Pathogens"},{"id":47664868,"name":"Biological sciences/Microbiology/Microbial communities/Microbial ecology"}],"tags":[],"updatedAt":"2025-12-03T08:09:18+00:00","versionOfRecord":{"articleIdentity":"rs-6079410","link":"https://doi.org/10.1038/s41467-025-66564-4","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2025-12-02 05:00:00","publishedOnDateReadable":"December 2nd, 2025"},"versionCreatedAt":"2025-04-29 08:24:20","video":"","vorDoi":"10.1038/s41467-025-66564-4","vorDoiUrl":"https://doi.org/10.1038/s41467-025-66564-4","workflowStages":[]},"version":"v1","identity":"rs-6079410","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6079410","identity":"rs-6079410","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}