The adaptive immune response in rheumatoid arthritis (RA) is influenced by an interaction between host genetics and environment, particularly the host microbiome. Association of the gut microbiota with various diseases has been reported, though the specific components of the microbiota that affect the host response leading to disease remain unknown. However, there is limited information on the role of gut microbiota in RA. In this study we aimed to define a microbial and metabolite profile that could predict disease status. In addition, we aimed to generate a humanized model of arthritis to confirm the RA-associated microbe. To identify an RA biomarker profile, the 16S ribosomal DNA of fecal samples from RA patients, first-degree relatives (to rule out environment/background as confounding factors), and random healthy non-RA controls were sequenced. Analysis of metabolites and their association with specific taxa was performed to investigate a potential mechanistic link. The role of an RA-associated microbe was confirmed using a human epithelial cell line and a humanized mouse model of arthritis. Patients with RA exhibited decreased gut microbial diversity compared with controls, which correlated with disease duration and autoantibody levels. A taxon-level analysis suggested an expansion of rare taxa, Actinobacteria, with a decrease in abundant taxa in patients with RA compared with controls. Prediction models based on the random forests algorithm suggested that three genera, Collinsella, Eggerthella, and Faecalibacterium, segregated with RA. The abundance of Collinsella correlated strongly with high levels of alpha-aminoadipic acid and asparagine as well as production of the proinflammatory cytokine IL-17A. A role for Collinsella in altering gut permeability and disease severity was confirmed in experimental arthritis. These observations suggest dysbiosis in RA patients resulting from the abundance of certain rare bacterial lineages. A correlation between the intestinal microbiota and metabolic signatures could determine a predictive profile for disease causation and progression.

Differential abundance analysis is at the core of statistical analysis of microbiome data. The compositional nature of microbiome sequencing data makes false positive control challenging. Here, we show that the compositional effects can be addressed by a simple, yet highly flexible and scalable, approach. The proposed method, LinDA, only requires fitting linear regression models on the centered log-ratio transformed data, and correcting the bias due to compositional effects. We show that LinDA enjoys asymptotic FDR control and can be extended to mixed-effect models for correlated microbiome data. Using simulations and real examples, we demonstrate the effectiveness of LinDA.

Endometrial cancer studies have led to a number of well-defined but mechanistically unconnected genetic and environmental risk factors. One of the emerging modulators between environmental triggers and genetic expression is the microbiome. We set out to inquire about the composition of the uterine microbiome and its putative role in endometrial cancer. We undertook a study of the microbiome in samples taken from different locations along the female reproductive tract in patients with endometrial cancer (n = 17), patients with endometrial hyperplasia (endometrial cancer precursor, n = 4), and patients afflicted with benign uterine conditions (n = 10). Vaginal, cervical, Fallopian, ovarian, peritoneal, and urine samples were collected aseptically both in the operating room and the pathology laboratory. DNA extraction was followed by amplification and high-throughput next generation sequencing (MiSeq) of the 16S rDNA V3-V5 region to identify the microbiota present. Microbiota data were summarized using both α-diversity to reflect species richness and evenness within bacterial populations and β-diversity to reflect the shared diversity between bacterial populations. Statistical significance was determined through the use of multiple testing, including the generalized mixed-effects model. The microbiome sequencing (16S rDNA V3-V5 region) revealed that the microbiomes of all organs (vagina, cervix, Fallopian tubes, and ovaries) are significantly correlated (p < 0.001) and that there is a structural microbiome shift in the cancer and hyperplasia cases, distinguishable from the benign cases (p = 0.01). Several taxa were found to be significantly enriched in samples belonging to the endometrial cancer cohort: Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1–68, Ruminococcus, and Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Of particular relevance, the simultaneous presence of Atopobium vaginae and an uncultured representative of the Porphyromonas sp. (99 % match to P. somerae) were found to be associated with disease status, especially if combined with a high vaginal pH (>4.5). Our results suggest that the detection of A. vaginae and the identified Porphyromonas sp. in the gynecologic tract combined with a high vaginal pH is statistically associated with the presence of endometrial cancer. Given the documented association of the identified microorganisms with other pathologies, these findings raise the possibility of a microbiome role in the manifestation, etiology, or progression of endometrial cancer that should be further investigated.

Background: Differential abundance analysis (DAA) is one central statistical task in microbiome data analysis. A robust and powerful DAA tool can help identify highly confident microbial candidates for further biological validation. Numerous DAA tools have been proposed in the past decade addressing the special characteristics of microbiome data such as zero inflation and compositional effects. Disturbingly, different DAA tools could sometimes produce quite discordant results, opening to the possibility of cherry-picking the tool in favor of one’s own hypothesis. To recommend the best DAA tool or practice to the field, a comprehensive evaluation, which covers as many biologically relevant scenarios as possible, is critically needed.Results: We performed by far the most comprehensive evaluation of existing DAA tools using real data-based simulations. We found that DAA methods explicitly addressing compositional effects such as ANCOM-BC, Aldex2, metagenomeSeq (fitFeatureModel), and DACOMP did have improved performance in false-positive control. But they are still not optimal: type 1 error inflation or low statistical power has been observed in many settings. The recent LDM method generally had the best power, but its false-positive control in the presence of strong compositional effects was not satisfactory. Overall, none of the evaluated methods is simultaneously robust, powerful, and flexible, which makes the selection of the best DAA tool difficult. To meet the analysis needs, we designed an optimized procedure, ZicoSeq, drawing on the strength of the existing DAA methods. We show that ZicoSeq generally controlled for false positives across settings, and the power was among the highest. Application of DAA methods to a large collection of real datasets revealed a similar pattern observed in simulation studies.Conclusions: Based on the benchmarking study, we conclude that none of the existing DAA methods evaluated can be applied blindly to any real microbiome dataset. The applicability of an existing DAA method depends on specific settings, which are usually unknown a priori. To circumvent the difficulty of selecting the best DAA tool in practice, we design ZicoSeq, which addresses the major challenges in DAA and remedies the drawbacks of existing DAA methods. ZicoSeq can be applied to microbiome datasets from diverse settings and is a useful DAA tool for robust microbiome biomarker discovery.

Normalization is the first critical step in microbiome sequencing data analysis used to account for variable library sizes. Current RNA-Seq based normalization methods that have been adapted for microbiome data fail to consider the unique characteristics of microbiome data, which contain a vast number of zeros due to the physical absence or under-sampling of the microbes. Normalization methods that specifically address the zero-inflation remain largely undeveloped. Here we propose geometric mean of pairwise ratios—a simple but effective normalization method—for zero-inflated sequencing data such as microbiome data. Simulation studies and real datasets analyses demonstrate that the proposed method is more robust than competing methods, leading to more powerful detection of differentially abundant taxa and higher reproducibility of the relative abundances of taxa.