High-throughput sequencing of PCR-amplified taxonomic markers (like the 16S rRNA gene) has enabled a new level of analysis of complex bacterial communities known as microbiomes. Many tools exist to quantify and compare abundance levels or OTU composition of communities in different conditions. The sequencing reads have to be denoised and assigned to the closest taxa from a reference database. Common approaches use a notion of 97% similarity and normalize the data by subsampling to equalize library sizes. In this paper, we show that statistical models allow more accurate abundance estimates. By providing a complete workflow in R, we enable the user to do sophisticated downstream statistical analyses, whether parametric or nonparametric. We provide examples of using the R packages dada2, phyloseq, DESeq2, ggplot2 and vegan to filter, visualize and test microbiome data. We also provide examples of supervised analyses using random forests and nonparametric testing using community networks and the ggnetwork package. Keywords Amendments from Version 1In version 2 of the manuscript:We have updated the procedure for storing the filtered and trimmed files during the call to dada2, this avoids overwriting the files if the workflow is run several times.We have replaced the msa alignment function with AlignSeqs function from from the DECIPHER 1 package, making the workflow more computationally efficient.We have expanded the phyloseq section and reduced the number of network plots. We have also provided detailed discussion of our choice not to make the PCoA and PCA plots square.We have added more detailed instructions in the Github repository as to how one can run only parts of the workflow and how to generate the full paper from scratch using the main.rnw file.As suggested by reviewers, we have added more extended captions to figures. We have however refrained from providing a complete evaluation of DADA2 vs. OTUs or pooled/unpooled data to this manuscript. Performing such evaluations well is a significant undertaking and would take significant space to explain, and our primary purpose here is to demonstrate the many features of an R/Bioconductor amplicon analysis workflow.We thank the three reviewers and a commentator who have provided useful feedback and we hope the revision has enhanced the readability and explained the code more completely.

Summary The indigenous microbiota of the nasal cavity plays important roles in human health and disease. Patterns of spatial variation in microbiota composition may help explain Staphylococcus aureus colonization, and reveal interspecies and species-host interactions. To assess the biogeography of the nasal microbiota, we sampled healthy subjects, representing both S. aureus carriers and non-carriers, at 3 nasal sites (anterior naris, middle meatus, and sphenoethmoidal recess). Phylogenetic compositional and sparse linear discriminant analyses revealed communities that differed according to site epithelium type and S. aureus culture-based carriage status. Corynebacterium accolens and C. pseudodiphtheriticum were identified as the most important microbial community determinants of S. aureus carriage, with competitive interactions evident only at sites with ciliated pseudostratified columnar epithelium. In vitro co-cultivation experiments provided supporting evidence of interactions among these species. These results highlight spatial variation in nasal microbial communities and differences in community composition between S. aureus carriers and non-carriers.

High-throughput sequencing of PCR-amplified taxonomic markers (like the 16S rRNA gene) has enabled a new level of analysis of complex bacterial communities known as microbiomes. Many tools exist to quantify and compare abundance levels or microbial composition of communities in different conditions. The sequencing reads have to be denoised and assigned to the closest taxa from a reference database. Common approaches use a notion of 97% similarity and normalize the data by subsampling to equalize library sizes. In this paper, we show that statistical models allow more accurate abundance estimates. By providing a complete workflow in R, we enable the user to do sophisticated downstream statistical analyses, including both parameteric and nonparametric methods. We provide examples of using the R packages dada2, phyloseq, DESeq2, ggplot2 and vegan to filter, visualize and test microbiome data. We also provide examples of supervised analyses using random forests, partial least squares and linear models as well as nonparametric testing using community networks and the ggnetwork package.

Innate natural killer (NK) cells are diverse at the single-cell level because of variegated expressions of activating and inhibitory receptors, yet the developmental roots and functional consequences of this diversity remain unknown. Because NK cells are critical for antiviral and antitumor responses, a better understanding of their diversity could lead to an improved ability to harness them therapeutically. We found that NK diversity is lower at birth than in adults. During an antiviral response to either HIV-1 or West Nile virus, NK diversity increases, resulting in terminal differentiation and cytokine production at the cost of cell division and degranulation. In African women matched for HIV-1 exposure risk, high NK diversity is associated with increased risk of HIV-1 acquisition. Existing diversity may therefore decrease the flexibility of the antiviral response. Collectively, the data reveal that human NK diversity is a previously undefined metric of immune history and function that may be clinically useful in forecasting the outcomes of infection and malignancy.

Although CD36 is generally recognized to be an inhibitory signaling receptor for thrombospondin-1 (TSP1), the molecular mechanism for transduction of this signal remains unclear. Based on evidence that myristic acid and TSP1 each modulate endothelial cell nitric oxide signaling in a CD36-dependent manner, we examined the ability of TSP1 to modulate the fatty acid translocase activity of CD36. TSP1 and a CD36 antibody that mimics the activity of TSP1 inhibited myristate uptake. Recombinant TSP1 type 1 repeats were weakly inhibitory, but an anti-angiogenic peptide derived from this domain potently inhibited myristate uptake. This peptide also inhibited membrane translocation of the myristoylated CD36 signaling target Fyn and activation of Src family kinases. Myristate uptake stimulated cGMP synthesis via endothelial nitric-oxide synthase and soluble guanylyl cyclase. CD36 ligands blocked myristate-stimulated cGMP accumulation in proportion to their ability to inhibit myristate uptake. TSP1 also inhibited myristate-stimulated cGMP synthesis by engaging its receptor CD47. Myristate stimulated endothelial and vascular smooth muscle cell adhesion on type I collagen via the NO/cGMP pathway, and CD36 ligands that inhibit myristate uptake blocked this response. Therefore, the fatty acid translocase activity of CD36 elicits proangiogenic signaling in vascular cells, and TSP1 inhibits this response by simultaneously inhibiting fatty acid uptake via CD36 and downstream cGMP signaling via CD47. Pathological angiogenesis or the lack thereof underlies a number of major diseases (1). Proangiogenic signals from vascular endothelial growth factors (VEGF)2 and fibroblast growth factors (FGF1 and FGF2) to induce blood vessel formation are opposed by signals from endogenous angiogenesis inhibitors, including two thrombospondins (TSP1 and TSP2) and proteolytic fragments of several extracellular matrix components (2, 3). Defining the mechanism of action of these inhibitors has been complicated by the finding that vascular cells express multiple receptors for several of these molecules. In the case of TSP1, endothelial cells express at least eight receptors, and some of these elicit pro-rather than anti-angiogenic responses (4, 5). The activities of some TSP1 receptors differ between large vessel and microvascular endothelial cells, and some are regulated by specific contextual signals (4 -6). CD36, a member of the scavenger receptor B family, is a TSP1 receptor that is selectively expressed in microvascular endothelium (7,8). CD36 was initially reported to recognize CSVTCG sequences in the type 1 repeats of TSP1 (9), but further studies identified higher affinity binding to the adjacent GVQXR sequences in the second and third type 1 repeats (10, 11). CD36 binding was markedly enhanced by epimerization of the first Ile in a peptide from the second type 1 repeat 434 GDGVITRIR 442 , where I (Ile) is the D isomer (11). TSP1, recombinant type 1 repeats of TSP1, and peptide mimetics of its CD36 binding sequences inhibit FGF2-stimulated endo...

Spatial and temporal patterns in microbial communities provide insights into the forces that shape them, their functions and roles in health and disease. Here, we used spatial and ecological statistics to analyze the role that saliva plays in structuring bacterial communities of the human mouth using >9000 dental and mucosal samples. We show that regardless of tissue type (teeth, alveolar mucosa, keratinized gingiva, or buccal mucosa), surface-associated bacterial communities vary along an ecological gradient from the front to the back of the mouth, and that on exposed tooth surfaces, the gradient is pronounced on lingual compared to buccal surfaces. Furthermore, our data suggest that this gradient is attenuated in individuals with low salivary flow due to Sjögren’s syndrome. Taken together, our findings imply that salivary flow influences the spatial organization of microbial communities and that biogeographical patterns may be useful for understanding host physiological processes and for predicting disease.

Pregnant women experience increased morbidity and mortality after influenza infection, for reasons that are not understood. Although some data suggest that natural killer (NK)-and T-cell responses are suppressed during pregnancy, influenza-specific responses have not been previously evaluated. Thus, we analyzed the responses of women that were pregnant (n = 21) versus those that were not (n = 29) immediately before inactivated influenza vaccination (IIV), 7 d after vaccination, and 6 wk postpartum. Expression of CD107a (a marker of cytolysis) and production of IFN-γ and macrophage inflammatory protein (MIP) 1β were assessed by flow cytometry. Pregnant women had a significantly increased percentage of NK cells producing a MIP-1β response to pH1N1 virus compared with nonpregnant women pre-IIV [median, 6.66 vs. 0.90% (P = 0.0149)] and 7 d post-IIV [median, 11.23 vs. 2.81% (P = 0.004)], indicating a heightened chemokine response in pregnant women that was further enhanced by the vaccination. Pregnant women also exhibited significantly increased T-cell production of MIP-1β and polyfunctionality in NK and T cells to pH1N1 virus pre-and post-IIV. NK-and T-cell polyfunctionality was also enhanced in pregnant women in response to the H3N2 viral strain. In contrast, pregnant women had significantly reduced NK-and T-cell responses to phorbol 12-myristate 13-acetate and ionomycin. This type of stimulation led to the conclusion that NK-and T-cell responses during pregnancy are suppressed, but clearly this conclusion is not correct relative to the more biologically relevant assays described here. Robust cellular immune responses to influenza during pregnancy could drive pulmonary inflammation, explaining increased morbidity and mortality.

Our work focuses on the stability, resilience, and response to perturbation of the bacterial communities in the human gut. Informative flash flood-like disturbances that eliminate most gastrointestinal biomass can be induced using a clinically-relevant iso-osmotic agent. We designed and executed such a disturbance in human volunteers using a dense longitudinal sampling scheme extending before and after induced diarrhea. This experiment has enabled a careful multidomain analysis of a controlled perturbation of the human gut microbiota with a new level of resolution. These new longitudinal multidomain data were analyzed using recently developed statistical methods that demonstrate improvements over current practices. By imposing sparsity constraints we have enhanced the interpretability of the analyses and by employing a new adaptive generalized principal components analysis, incorporated modulated phylogenetic information and enhanced interpretation through scoring of the portions of the tree most influenced by the perturbation. Our analyses leverage the taxa-sample duality in the data to show how the gut microbiota recovers following this perturbation. Through a holistic approach that integrates phylogenetic, metagenomic and abundance information, we elucidate patterns of taxonomic and functional change that characterize the community recovery process across individuals. We provide complete code and illustrations of new sparse statistical methods for high-dimensional, longitudinal multidomain data that provide greater interpretability than existing methods.