Microbiome Compositional Data Research Articles

Oral bacteria relative abundance in faeces increases due to gut microbiota depletion and is linked with patient outcomes.

The detection of oral bacteria in faecal samples has been associated with inflammation and intestinal diseases. The increased relative abundance of oral bacteria in faeces has two competing explanations: either oral bacteria invade the gut ecosystem and expand (the 'expansion' hypothesis), or oral bacteria transit through the gut and their relative increase marks the depletion of other gut bacteria (the 'marker' hypothesis). Here we collected oral and faecal samples from mouse models of gut dysbiosis (antibiotic treatment and DSS-induced colitis) and used 16S ribosomal RNA sequencing to determine the abundance dynamics of oral bacteria. We found that the relative, but not absolute, abundance of oral bacteria increases, reflecting the 'marker' hypothesis. Faecal microbiome datasets from diverse patient cohorts, including healthy individuals and patients with allogeneic haematopoietic cell transplantation or inflammatory bowel disease, consistently support the 'marker' hypothesis and explain associations between oral bacterial abundance and patient outcomes consistent with depleted gut microbiota. By distinguishing between the two hypotheses, our study guides the interpretation of microbiome compositional data and could potentially identify cases where therapies are needed to rebuild the resident microbiome rather than protect against invading oral bacteria.

Microbiome compositional data analysis for survival studies.

The growing interest in studying the relationship between the human microbiome and our health has also extended to time-to-event studies where researchers explore the connection between the microbiome and the occurrence of a specific event of interest. The analysis of microbiome obtained through high throughput sequencing techniques requires the use of specialized Compositional Data Analysis (CoDA) methods designed to accommodate its compositional nature. There is a limited availability of statistical tools for microbiome analysis that incorporate CoDA, and this is even more pronounced in the context of survival analysis. To fill this methodological gap, we present coda4microbiome for survival studies, a new methodology for the identification of microbial signatures in time-to-event studies. The algorithm implements an elastic-net penalized Cox regression model adapted to compositional covariates. We illustrate coda4microbiome algorithm for survival studies with a case study about the time to develop type 1 diabetes for non-obese diabetic mice. Our algorithm identified a bacterial signature composed of 21 genera associated with diabetes development. coda4microbiome for survival studies is integrated in the R package coda4microbiome as an extension of the existing functions for cross-sectional and longitudinal studies.

Compositional variable selection in quantile regression for microbiome data with false discovery rate control

AbstractAdvancement in high‐throughput sequencing technologies has stimulated intensive research interests to identify specific microbial taxa that are associated with disease conditions. Such knowledge is invaluable both from the perspective of understanding biology and from the biomedical perspective of therapeutic development, as the microbiome is inherently modifiable. Despite availability of massive data, analysis of microbiome compositional data remains difficult. The nature that relative abundances of all components of a microbial community sum to one poses challenges for statistical analysis, especially in high‐dimensional settings, where a common research theme is to select a small fraction of signals from amid many noisy features. Motivated by studies examining the role of microbiome in host transcriptomics, we propose a novel approach to identify microbial taxa that are associated with host gene expressions. Besides accommodating compositional nature of microbiome data, our method both achieves FDR‐controlled variable selection, and captures heterogeneity due to either heteroscedastic variance or non‐location‐scale covariate effects displayed in the motivating dataset. We demonstrate the superior performance of our method using extensive numerical simulation studies and then apply it to real‐world microbiome data analysis to gain novel biological insights that are missed by traditional mean‐based linear regression analysis.

Assessing the causal association between sleep apnea and the human gut microbiome composition: A two-sample Mendelian randomization study.

Studies have linked gut microbiota dysbiosis with sleep apnea; however, no causal relationship was found in human subjects. Finding new targets for the pathophysiology of sleep apnea might be made possible by systematically investigating the causal relationship between the human gut microbiota and sleep apnea. A two-sample Mendelian randomization analysis was conducted. The human gut microbiome composition data, spanning five taxonomic levels, were acquired from a genome-wide association study that included 18,340 participants from 24 cohorts. Genome-wide association study data for sleep apnea were obtained from the Sleep Disorder Knowledge Portal for primary analysis and the FinnGen consortium for meta-analysis. Sensitivity analyses were conducted to evaluate heterogeneity and pleiotropy. Using inverse-variance weighted analysis, eight microbial taxa were initially found to be substantially linked with the apnea-hypopnea index. Only three microbial taxa remained significant associations with sleep apnea when combined with the FinnGen consortium (the class Bacilli: B = 8.21%, 95% CI = 0.93%-15.49%; p = 0.03; the order Lactobacillales: B = 7.55%, 95% CI = 0.25%-4.85%; p = 0.04; the genus RuminococcaceaeUCG009: B = -21.63%, 95% CI = -41.47% to -1.80%; p = 0.03). Sleep apnea may lead to gut dysbiosis as significant reductions in butyrate-producing bacteria and increases in lactate-producing bacteria. By integrating genomes and metabolism, the evidence that three microbiome species are causally linked to sleep apnea may offer a fresh perspective on the underlying mechanisms of the condition.

Host genomic variation shapes gut microbiome diversity in threespine stickleback fish.

A major focus of host-microbe research is to understand how genetic differences, of various magnitudes, among hosts translate to differences in their microbiomes. This has been challenging for animal hosts, including humans, because it is difficult to control environmental variables tightly enough to isolate direct genetic effects on the microbiome. Our work in stickleback fish is a significant contribution because our experimental approach allowed strict control over environmental factors, including standardization of the microbiome from the earliest stage of development and unrestricted co-housing of fish in a truly common environment. Furthermore, we measured host genetic variation over 2,000 regions of the stickleback genome, comparing this information and microbiome composition data among fish from very similar and very different genetic backgrounds. Our findings highlight how differences in the host genome influence microbiome diversity and make a case for future manipulative microbiome experiments that use host systems with naturally occurring genetic variation.

Dependence Structure Analysis and Its Application in Human Microbiome

The human microbiome has been recently shown to be associated with disease risks and has important implications in risk stratification and precision medicine. Due to abundant taxa in the human body, microbiome data are high-dimensional and compositional. Dirichlet distributions and their generalization are used to characterize the dependence structures of microbial data. Another existing method for fitting microbiome data employed Gaussian graphical model using the centered log-transformation (CLR). However, Dirichlet distributions are not able to infer networks or to estimate some extremely rare probabilities. On the other hand, it is hard to interpret the network analysis results using CLR. Furthermore, the data analysis showed that there is a lack of efficient multivariate distributions for fitting microbiome data, which results in inadequate statistical inferences. In this paper, we propose new multivariate distributions for modeling the dependence structures of the high dimensional and compositional microbiome data using inverse gamma distributions and copula techniques. The data analysis in the American gut project shows our proposed methods perform well.

CRAMdb: a comprehensive database for composition and roles of microbiome in animals.

CRAMdb (a database for composition and roles of animal microbiome) is a comprehensive resource of curated and consistently annotated metagenomes for non-human animals. It focuses on the composition and roles of the microbiome in various animal species. The main goal of the CRAMdb is to facilitate the reuse of animal metagenomic data, and enable cross-host and cross-phenotype comparisons. To this end, we consistently annotated microbiomes (including 16S, 18S, ITSand metagenomics sequencing data) of 516 animals from 475 projects spanning 43 phenotype pairs to construct the database that is equipped with 9430 bacteria, 278 archaea, 2216 fungiand 458 viruses. CRAMdb provides two main contents: microbiome composition data, illustrating the landscape of the microbiota (bacteria, archaea, fungi, and viruses) in various animal species, and microbiome association data, revealing the relationships between the microbiota and various phenotypes across different animal species. More importantly, users can quickly compare the composition of the microbiota of interest cross-host or body site and the associated taxa that differ between phenotype pairs cross-host or cross-phenotype. CRAMdb is freely available at (http://www.ehbio.com/CRAMdb).

BREAst Cancer Personalised NuTrition (BREACPNT): dietary intervention in breast cancer survivors treated with endocrine therapy – a protocol for a randomised clinical trial

IntroductionBreast cancer survivors treated with adjuvant endocrine therapy commonly experience weight gain, which has been associated with low adherence to therapy and worse breast cancer prognosis. We aim to assess...

Association of plasma microbial composition with a leaky gut in obesity-related osteoarthritis: An exploratory study

ObjectiveTo examine the plasma microbiome for differences between obese individuals with and without osteoarthritis (OA) and its association with serum lipopolysaccharide (LPS). DesignBlood samples from 70 participants with body mass index (BMI) ​≥ ​30kg/m2 and age ≥55 years, with (cases) or without (controls) hand plus knee OA, were analyzed for serum LPS and composition of the plasma microbiome. The Dirichlet-multinominal recursive partitioning model (DM-RPart) was applied to microbiome compositional data to test the hypothesis that LPS levels distinguish plasma microbiome, accounting for BMI and age. ResultsNo significant differences in alpha diversity, or compositional differences between groups at the genus level, were seen between cases and controls (p ​= ​0.11). β-Diversity was significantly associated with serum LPS levels (p ​= ​0.01). DM-RPart resulted in an optimal tree with 3 divisions: 1) based on age (split at 69 years); 2) those older than 69 were split based on BMI; 3) those with BMI <39 ​kg/m2 were split based on LPS level (at 65 EU/ml). This resulted in 4 groups (nodes 2, and 5–7). Participants in node 2 were younger and the majority had no or mild OA. Those in nodes 5 and 6 were comparable in age and BMI but node 6 had higher LPS and more severe OA. Individuals in node 7 were older, had higher BMI, and the most severe OA. ConclusionsOur results suggest a relationship between serum LPS and the plasma microbiome in a subgroup of obese individuals with hand plus knee OA that could reflect differences in intestinal permeability.

LOCOM: A logistic regression model for testing differential abundance in compositional microbiome data with false discovery rate control

Compositional analysis is based on the premise that a relatively small proportion of taxa are differentially abundant, while the ratios of the relative abundances of the remaining taxa remain unchanged. Most existing methods use log-transformed data, but log-transformation of data with pervasive zero counts is problematic, and these methods cannot always control the false discovery rate (FDR). Further, high-throughput microbiome data such as 16S amplicon or metagenomic sequencing are subject to experimental biases that are introduced in every step of the experimental workflow. McLaren et al. [eLife 8, e46923 (2019)] have recently proposed a model for how these biases affect relative abundance data. Motivated by this model, we show that the odds ratios in a logistic regression comparing counts in two taxa are invariant to experimental biases. With this motivation, we propose logistic compositional analysis (LOCOM), a robust logistic regression approach to compositional analysis, that does not require pseudocounts. Inference is based on permutation to account for overdispersion and small sample sizes. Traits can be either binary or continuous, and adjustment for confounders is supported. Our simulations indicate that LOCOM always preserved FDR and had much improved sensitivity over existing methods. In contrast, analysis of composition of microbiomes (ANCOM) and ANCOM with bias correction (ANCOM-BC)/ANOVA-Like Differential Expression tool (ALDEx2) had inflated FDR when the effect sizes were small and large, respectively. Only LOCOM was robust to experimental biases in every situation. The flexibility of our method for a variety of microbiome studies is illustrated by the analysis of data from two microbiome studies. Our R package LOCOM is publicly available.

BiomeHorizon: Visualizing Microbiome Time Series Data in R.

ABSTRACTDespite playing a key role in the health of their hosts, host-associated microbial communities demonstrate considerable variation over time. These communities comprise thousands of temporally dynamic taxa, which makes visualizing microbial time series data challenging. As such, a method to visualize both the proportional and absolute change in the relative abundance of multiple taxa across multiple subjects over time is needed. To address this gap, we developed BiomeHorizon, the first automated, open-source R package that visualizes longitudinal compositional microbiome data using horizon plots. BiomeHorizon is available at https://github.com/blekhmanlab/biomehorizon/ and a guide with step-by-step instructions for using the package is provided at https://blekhmanlab.github.io/biomehorizon/.IMPORTANCE Host-associated microbiota (i.e., the number and types of bacteria in the body) can have profound impacts on an animal’s day-to-day functioning as well as their long-term health. Recent work has shown that these microbial communities change substantially over time, so it is important to be able to link changes in the abundance of certain microbes with host health outcomes. However, visualizing such changes is difficult because the microbiome comprises thousands of different microbes. To address this issue, we developed BiomeHorizon, an R package for visualizing longitudinal microbiome data using horizon plots. BiomeHorizon accepts a range of data formats and was developed with two common microbiome study designs in mind: human health studies, where the microbiome is sampled at set time points, and observational wildlife studies, where samples may be collected at irregular time intervals. BiomeHorizon thus provides a flexible, user-friendly approach to microbiome time series data visualization and analysis.

LinDA: linear models for differential abundance analysis of microbiome compositional data

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.

Dietary carbohydrate intake is associated with the subgingival plaque oral microbiome abundance and diversity in a cohort of postmenopausal women

Limited research exists on carbohydrate intake and oral microbiome diversity and composition assessed with next-generation sequencing. We aimed to better understand the association between habitual carbohydrate intake and the oral microbiome, as the oral microbiome has been associated with caries, periodontal disease, and systemic diseases. We investigated if total carbohydrates, starch, monosaccharides, disaccharides, fiber, or glycemic load (GL) were associated with the diversity and composition of oral bacteria in subgingival plaque samples of 1204 post-menopausal women. Carbohydrate intake and GL were assessed from a food frequency questionnaire, and adjusted for energy intake. The V3–V4 region of the 16S rRNA gene from subgingival plaque samples were sequenced to identify the relative abundance of microbiome compositional data expressed as operational taxonomic units (OTUs). The abundance of OTUs were centered log(2)-ratio transformed to account for the compositional data structure. Associations between carbohydrate/GL intake and microbiome alpha-diversity measures were examined using linear regression. PERMANOVA analyses were conducted to examine microbiome beta-diversity measures across quartiles of carbohydrate/GL intake. Associations between intake of carbohydrates and GL and the abundance of the 245 identified OTUs were examined by using linear regression. Total carbohydrates, GL, starch, lactose, and sucrose intake were inversely associated with alpha-diversity measures. Beta-diversity across quartiles of total carbohydrates, fiber, GL, sucrose, and galactose, were all statistically significant (p for PERMANOVA p < 0.05). Positive associations were observed between total carbohydrates, GL, sucrose and Streptococcus mutans; GL and both Sphingomonas HOT 006 and Scardovia wiggsiae; and sucrose and Streptococcus lactarius. A negative association was observed between lactose and Aggregatibacter segnis, and between sucrose and both TM7_[G-1] HOT 346 and Leptotrichia HOT 223. Intake of total carbohydrate, GL, and sucrose were inversely associated with subgingival bacteria alpha-diversity, the microbial beta-diversity varied by their intake, and they were associated with the relative abundance of specific OTUs. Higher intake of sucrose, or high GL foods, may influence poor oral health outcomes (and perhaps systemic health outcomes) in older women via their influence on the oral microbiome.

Analysis of gut microbiota in patients with epilepsy treated with valproate: Results from a three months observational prospective cohort study

BackgroundGrowing evidence implicates the potential effect of microbiota on the pathogenesis and course of epilepsy. However, the effects of valproate (VPA), a broad spectrum anti-epileptic drugs, on gut microbiota have not been investigated in humans. This study aimed to analyze fecal microbiota in patients with epilepsy treated with valproate. MethodsA total of 10 participants, who were newly diagnosed of cryptogenic epilepsy with treatment naïve and received 1000 mg daily doses of VPA, were recruited in our prospective study. Microbiota compositions were evaluated at baseline and after three months of VPA treatment using 16S rDNA sequencing. ResultsVPA treatment was associated with clinical improvements in all patients, but not changes in gut microbiota richness and complexity (Shannon: p = 0.82). Microbiome composition structure differences also revealed no statistical difference in dissimilarity (Adonis: p = 0.90). No statistical difference taxa were found between two groups. However, the ratio of phyla Firmicutes to Bacteriodetes (ANOVA: p = 0.037) markedly raised after three months of VPA-treatment. A correlation matrix based on the spearman correlation distance confirmed associations between specific fecal taxa and VPA-related clinical metabolic parameters, including drug concentration in the blood, total cholesterol, triglyceride, lactate dehydrogenase, alanine aminotransferase, aspartate aminotransferase and weight gain. (p < 0.05) ConclusionsAmong those patients treated with VPA, characterization of the gut microbiota altered, and gut microbiota associated with weight gain and clinical biochemical indexes, suggesting that microbiome composition data might involve in the mechanisms of VPA induced metabolic disorder.

Assessing the Reproducibility of Microbiome Measurements Based on Concordance Correlation Coefficients.

Evaluating the reproducibility or agreement of microbiome measurements is often a crucial step to ensure rigorous downstream analyses in microbiome studies. In this paper, we address this need by developing adaptations of Lin's concordance correlation coefficient (CCC) tailored to microbiome studies. We introduce a general formulation of the new CCC measures upon the use of a distance function appropriately characterizing the discrepancy between microbiome compositional measurements. We thoroughly study the special cases that adopt Euclidean distance and Aitchison distance. Our proposals appropriately account for the unique features of microbiome compositional data, including high-dimensionality, dependency among individual relative abundances, and the presence of many zeros. We further investigate a practical compound approach to help better understand the sources of data inconsistency. Extensive simulation studies are conducted to evaluate the utility of the proposed methods in realistic scenarios. We also apply the proposed methods to a microbiome validation dataset from the Feeding Infants Right.. from the STart (FIRST) study. Our analyses offer useful insight about the extent of data variations resulted from two different experiment procedures as well as their heterogeneous patterns across genera.

Assessing mediating effects of high-dimensional microbiome measurements in dietary intervention studies.

Habitual diet can influence health-related outcomes directly, but such effects may also be modulated indirectly by gut microbiota. We consider randomized trials and the question to what extent the effect of diet on an outcome of interest is mediated through the gut microbiome or whether there is a diet-microbiome interaction identifying subgroups of individuals who are more susceptible to specific dietary effects. The baseline microbiome by itself may be a modifier of the effects of diet on health. Yet, the high dimensionality of microbiome data requires innovative statistical approaches to identify potential mediating or moderating effects. To motivate our proposal for an appropriate analysis workflow, we consider a randomized trial that investigates the effect of a 4-week vegan diet on the diversity of gut microbiota and branched-chain amino acid metabolism in healthy omnivorous volunteers. To address the challenge of compositional microbiome data, we consider an adaptation of the lasso for penalized estimation of multivariable regression models with a large number of microbiotic taxa. This is plugged into a classical regression mediation effect analysis strategy. The interaction effects are obtained via an approach that can directly estimate them without having to deal with main effects. As a result we obtain signatures comprised of microbiotic taxa with potential mediating and moderating effects. Some taxa no longer show up as mediating, when taking moderating effects into account. Thus, the proposed analysis strategy allows to identify specific mediating effects, while avoiding potential erroneous conclusions, where moderating effects might have believed to be mediatingeffects.

Log-ratio analysis of microbiome data with many zeroes is library size dependent.

Microbiome composition data collected through amplicon sequencing are count data on taxa in which the total count per sample (the library size) is an artefact of the sequencing platform, and as a result, such data are compositional. To avoid library size dependency, one common way of analysing multivariate compositional data is to perform a principal component analysis (PCA) on data transformed with the centred log‐ratio, hereafter called a log‐ratio PCA. Two aspects typical of amplicon sequencing data are the large differences in library size and the large number of zeroes. In this study, we show on real data and by simulation that, applied to data that combine these two aspects, log‐ratio PCA is nevertheless heavily dependent on the library size. This leads to a reduction in power when testing against any explanatory variable in log‐ratio redundancy analysis. If there is additionally a correlation between the library size and the explanatory variable, then the type 1 error becomes inflated. We explore putative solutions to this problem.

Modeling Longitudinal Microbiome Compositional Data: A Two-Part Linear Mixed Model with Shared Random Effects

Longitudinal microbiome studies have been widely used to unveil the dynamics in the complex host-microbial ecosystems. Modeling the longitudinal microbiome compositional data, which is semi-continuous in nature, is challenging in several aspects: the overabundance of zeros, the heavy skewness of non-zero values that are bounded in (0, 1), and the dependence between the binary and non-zero parts. To deal with these challenges, we first extended the work of Chen and Li [1] and proposed a two-part zero-inflated Beta regression model with shared random effects (ZIBR-SRE), which characterize the dependence between the binary and the continuous parts. Besides, the microbiome compositional data have unit-sum constraint, indicating the existence of negative correlations among taxa. As ZIBR-SRE models each taxon separately, it does not satisfy the sum-to-one constraint. We then proposed a two-part linear mixed model (TPLMM) with shared random effects to formulate the log-transformed standardized relative abundances rather than the original ones. Such transformation is called “additive logistic transformation”, initially developed for cross-sectional compositional data. We extended it to analyze the longitudinal microbiome compositions and showed that the unit-sum constraint can be automatically satisfied under the TPLMM framework. Model performances of TPLMM and ZIBR-SRE were compared with existing methods in simulation studies. Under settings adopted from real data, TPLMM had the best performance and is recommended for practical use. An oral microbiome application further showed that TPLMM and ZIBR-SRE estimated a strong correlation structure in the binary and the continuous parts, suggesting models without accounting for this dependence would lead to biased inferences.

A t-SNE Based Classification Approach to Compositional Microbiome Data.

As a data-driven dimensionality reduction and visualization tool, t-distributed stochastic neighborhood embedding (t-SNE) has been successfully applied to a variety of fields. In recent years, it has also received increasing attention for classification and regression analysis. This study presented a t-SNE based classification approach for compositional microbiome data, which enabled us to build classifiers and classify new samples in the reduced dimensional space produced by t-SNE. The Aitchison distance was employed to modify the conditional probabilities in t-SNE to account for the compositionality of microbiome data. To classify a new sample, its low-dimensional features were obtained as the weighted mean vector of its nearest neighbors in the training set. Using the low-dimensional features as input, three commonly used machine learning algorithms, logistic regression (LR), support vector machine (SVM), and decision tree (DT) were considered for classification tasks in this study. The proposed approach was applied to two disease-associated microbiome datasets, achieving better classification performance compared with the classifiers built in the original high-dimensional space. The analytic results also showed that t-SNE with Aitchison distance led to improvement of classification accuracy in both datasets. In conclusion, we have developed a t-SNE based classification approach that is suitable for compositional microbiome data and may also serve as a baseline for more complex classification models.

Mediation effect selection in high-dimensional and compositional microbiome data.

The microbiome plays an important role in human health by mediating the path from environmental exposures to health outcomes. The relative abundances of the high-dimensional microbiome data have an unit-sum restriction, rendering standard statistical methods in the Euclidean space invalid. To address this problem, we use the isometric log-ratio transformations of the relative abundances as the mediator variables. To select significant mediators, we consider a closed testing-based selection procedure with desirable confidence. Simulations are provided to verify the effectiveness of our method. As an illustrative example, we apply the proposed method to study the mediation effects of murine gut microbiome between subtherapeutic antibiotic treatment and body weight gain, and identify Coprobacillus and Adlercreutzia as two significant mediators.