Managing batch effects in microbiome data

The salivary microbiome as a diagnostic biomarker of periodontitis: a 16S multi-batch study before and after the removal of batch effects

Introduction

Microbiome-based clinical applications that improve diagnosis related to oral health are of great interest to precision dentistry. Predictive studies on the salivary microbiome are scarce and of low methodological quality (low sample sizes, lack of biological heterogeneity, and absence of a validation process). None of them evaluates the impact of confounding factors as batch effects (BEs). This is the first 16S multi-batch study to analyze the salivary microbiome at the amplicon sequence variant (ASV) level in terms of differential abundance and machine learning models. This is done in periodontally healthy and periodontitis patients before and after removing BEs.

Methods

Saliva was collected from 124 patients (50 healthy, 74 periodontitis) in our setting. Sequencing of the V3-V4 16S rRNA gene region was performed in Illumina MiSeq. In parallel, searches were conducted on four databases to identify previous Illumina V3-V4 sequencing studies on the salivary microbiome. Investigations that met predefined criteria were included in the analysis, and the own and external sequences were processed using the same bioinformatics protocol. The statistical analysis was performed in the R-Bioconductor environment.

Results

The elimination of BEs reduced the number of ASVs with differential abundance between the groups by approximately one-third (Before=265; After=190). Before removing BEs, the model constructed using all study samples (796) comprised 16 ASVs (0.16%) and had an area under the curve (AUC) of 0.944, sensitivity of 90.73%, and specificity of 87.16%. The model built using two-thirds of the specimens (training=531) comprised 35 ASVs (0.36%) and had an AUC of 0.955, sensitivity of 86.54%, and specificity of 90.06% after being validated in the remaining one-third (test=265). After removing BEs, the models required more ASVs (all samples=200-2.03%; training=100-1.01%) to obtain slightly lower AUC (all=0.935; test=0.947), lower sensitivity (all=81.79%; test=78.85%), and similar specificity (all=91.51%; test=90.68%).

Conclusions

The removal of BEs controls false positive ASVs in the differential abundance analysis. However, their elimination implies a significantly larger number of predictor taxa to achieve optimal performance, creating less robust classifiers. As all the provided models can accurately discriminate health from periodontitis, implying good/excellent sensitivities/specificities, the salivary microbiome demonstrates potential clinical applicability as a precision diagnostic tool for periodontitis.

High-throughput DNA extraction strategy for fecal microbiome studies

Microbiome studies are becoming larger in size to detect the potentially small effect that environmental factors have on our gut microbiomes, or that the microbiome has on our health. Therefore, fast and reproducible DNA isolation methods are needed to handle thousands of fecal samples. We used the Chemagic 360 chemistry and Magnetic Separation Module I (MSMI) instrument to compare two sample preservatives and four different pre-treatment protocols to find an optimal method for DNA isolation from thousands of fecal samples. The pre-treatments included bead beating, sample handling in tube and plate format, and proteinase K incubation. The optimal method offers a sufficient yield of high-quality DNA without contamination. Three human fecal samples (adult, senior, and infant) with technical replicates were extracted. The extraction included negative controls (OMNIgeneGUT, DNA/RNA shield fluid, and Chemagic Lysis Buffer 1) to detect cross-contamination and ZymoBIOMICS Gut Microbiome Standard as a positive control to mimic the human gut microbiome and assess sensitivity of the extraction method. All samples were extracted using Chemagic DNA Stool 200 H96 kit (PerkinElmer, Finland). The samples were collected in two preservatives, OMNIgeneGUT and DNA/RNA shield fluid. DNA quantity was measured using Qubit-fluorometer, DNA purity and quality using gel electrophoresis, and taxonomic signatures with 16S rRNA gene-based sequencing with V3V4 and V4 regions. Bead beating increased bacterial diversity. The largest increase was detected in gram-positive genera Blautia, Bifidobacterium, and Ruminococcus. Preservatives showed minor differences in bacterial abundances. The profiles between the V3V4 and V4 regions differed considerably with lower diversity samples. Negative controls showed signs from genera abundant in fecal samples. Technical replicates of the Gut Standard and stool samples showed low variation. The selected isolation protocol included recommended steps from manufacturer as well as bead beating. Bead beating was found to be necessary to detect hard-to-lyse bacteria. The protocol was reproducible in terms of DNA yield among different stool replicates and the ZymoBIOMICS Gut Microbiome Standard. The MSM1 instrument and pre-treatment in a 96-format offered the possibility of automation and handling of large sample collections. Both preservatives were feasible in terms of sample handling and had low variation in taxonomic signatures. The 16S rRNA target region had a high impact on the composition of the bacterial profile.

IMPORTANCE

Next-generation sequencing (NGS) is a widely used method for determining the composition of the gut microbiota. Due to the differences in the gut microbiota composition between individuals, microbiome studies have expanded into large population studies to maximize detection of small effects on microbe-host interactions. Thus, the demand for a rapid and reliable microbial profiling is continuously increasing, making the optimization of high-throughput 96-format DNA extraction integral for NGS-based downstream applications. However, experimental protocols are prone to bias and errors from sample collection and storage, to DNA extraction, primer selection and sequencing, and bioinformatics analyses. Methodological bias can contribute to differences in microbiome profiles, causing variability across studies and laboratories using different protocols. To improve consistency and confidence of the measurements, the standardization of microbiome analysis methods has been recognized in many fields.

Identification and validation of microbial biomarkers from cross-cohort datasets using xMarkerFinder

Microbial signatures have emerged as promising biomarkers for disease diagnostics and prognostics, yet their variability across different studies calls for a standardized approach to biomarker research. Therefore, we introduce xMarkerFinder, a four-stage computational framework for microbial biomarker identification with comprehensive validations from cross-cohort datasets, including differential signature identification, model construction, model validation and biomarker interpretation. xMarkerFinder enables the identification and validation of reproducible biomarkers for cross-cohort studies, along with the establishment of classification models and potential microbiome-induced mechanisms. Originally developed for gut microbiome research, xMarkerFinder's adaptable design makes it applicable to various microbial habitats and data types. Distinct from existing biomarker research tools that typically concentrate on a singular aspect, xMarkerFinder uniquely incorporates a sophisticated feature selection process, specifically designed to address the heterogeneity between different cohorts, extensive internal and external validations, and detailed specificity assessments. Execution time varies depending on the sample size, selected algorithm and computational resource. Accessible via GitHub ( https://github.com/tjcadd2020/xMarkerFinder ), xMarkerFinder supports users with diverse expertise levels through different execution options, including step-to-step scripts with detailed tutorials and frequently asked questions, a single-command execution script, a ready-to-use Docker image and a user-friendly web server ( https://www.biosino.org/xmarkerfinder ).

The gut microbiomes of Channel Island foxes and island spotted skunks exhibit fine-scale differentiation across host species and island populations

California's Channel Islands are home to two endemic mammalian carnivores: island foxes (Urocyon littoralis) and island spotted skunks (Spilogale gracilis amphiala). Although it is rare for two insular terrestrial carnivores to coexist, these known competitors persist on both Santa Cruz Island and Santa Rosa Island. We hypothesized that examination of their gut microbial communities would provide insight into the factors that enable this coexistence, as microbial symbionts often reflect host evolutionary history and contemporary ecology. Using rectal swabs collected from island foxes and island spotted skunks sampled across both islands, we generated 16S rRNA amplicon sequencing data to characterize their gut microbiomes. While island foxes and island spotted skunks both harbored the core mammalian microbiome, host species explained the largest proportion of variation in the dataset. We further identified intraspecific variation between island populations, with greater differentiation observed between more specialist island spotted skunk populations compared to more generalist island fox populations. This pattern may reflect differences in resource utilization following fine-scale niche differentiation. It may further reflect evolutionary differences regarding the timing of intraspecific separation. Considered together, this study contributes to the growing catalog of wildlife microbiome studies, with important implications for understanding how eco-evolutionary processes enable the coexistence of terrestrial carnivores-and their microbiomes-in island environments.

Microbial dysbiosis index for assessing colitis status in mouse models: A systematic review and meta-analysis

Although countless gut microbiome studies on colitis using mouse models have been carried out, experiments with small sample sizes have encountered reproducibility limitations because of batch effects and statistical errors. In this study, dextran-sodium-sulfate-induced microbial dysbiosis index (DiMDI) was introduced as a reliable dysbiosis index that can be used to assess the state of microbial dysbiosis in DSS-induced mouse models. Meta-analysis of 189 datasets from 11 independent studies was performed to construct the DiMDI. Microbial dysbiosis biomarkers, Muribaculaceae, Alistipes, Turicibacter, and Bacteroides, were selected through four different feature selection methods and used to construct the DiMDI. This index demonstrated a high accuracy of 82.3% and showed strong robustness (88.9%) in the independent cohort. Therefore, DiMDI may be used as a standard for assessing microbial imbalance in DSS-induced mouse models and may contribute to the development of reliable colitis microbiome studies in mouse experiments.

Methodology for biomarker discovery with reproducibility in microbiome data using machine learning

BACKGROUND

In recent years, human microbiome studies have received increasing attention as this field is considered a potential source for clinical applications. With the advancements in omics technologies and AI, research focused on the discovery for potential biomarkers in the human microbiome using machine learning tools has produced positive outcomes. Despite the promising results, several issues can still be found in these studies such as datasets with small number of samples, inconsistent results, lack of uniform processing and methodologies, and other additional factors lead to lack of reproducibility in biomedical research. In this work, we propose a methodology that combines the DADA2 pipeline for 16s rRNA sequences processing and the Recursive Ensemble Feature Selection (REFS) in multiple datasets to increase reproducibility and obtain robust and reliable results in biomedical research.

RESULTS

Three experiments were performed analyzing microbiome data from patients/cases in Inflammatory Bowel Disease (IBD), Autism Spectrum Disorder (ASD), and Type 2 Diabetes (T2D). In each experiment, we found a biomarker signature in one dataset and applied to 2 other as further validation. The effectiveness of the proposed methodology was compared with other feature selection methods such as K-Best with F-score and random selection as a base line. The Area Under the Curve (AUC) was employed as a measure of diagnostic accuracy and used as a metric for comparing the results of the proposed methodology with other feature selection methods. Additionally, we use the Matthews Correlation Coefficient (MCC) as a metric to evaluate the performance of the methodology as well as for comparison with other feature selection methods.

CONCLUSIONS

We developed a methodology for reproducible biomarker discovery for 16s rRNA microbiome sequence analysis, addressing the issues related with data dimensionality, inconsistent results and validation across independent datasets. The findings from the three experiments, across 9 different datasets, show that the proposed methodology achieved higher accuracy compared to other feature selection methods. This methodology is a first approach to increase reproducibility, to provide robust and reliable results.

Comprehensive assessment of machine learning methods for diagnosing gastrointestinal diseases through whole metagenome sequencing data

The gut microbiome, linked significantly to host diseases, offers potential for disease diagnosis through machine learning (ML) pipelines. These pipelines, crucial in modeling diseases using high-dimensional microbiome data, involve selecting profile modalities, data preprocessing techniques, and classification algorithms, each impacting the model accuracy and generalizability. Despite whole metagenome shotgun sequencing (WMS) gaining popularity for human gut microbiome profiling, a consensus on the optimal methods for ML pipelines in disease diagnosis using WMS data remains elusive. Addressing this gap, we comprehensively evaluated ML methods for diagnosing Crohn's disease and colorectal cancer, using 2,553 fecal WMS samples from 21 case-control studies. Our study uncovered crucial insights: gut-specific, species-level taxonomic features proved to be the most effective for profiling; batch correction was not consistently beneficial for model performance; compositional data transformations markedly improved the models; and while nonlinear ensemble classification algorithms typically offered superior performance, linear models with proper regularization were found to be more effective for diseases that are linearly separable based on microbiome data. An optimal ML pipeline, integrating the most effective methods, was validated for generalizability using holdout data. This research offers practical guidelines for constructing reliable disease diagnostic ML models with fecal WMS data.

Multi-omic approaches for host-microbiome data integration

The gut microbiome interacts with the host through complex networks that affect physiology and health outcomes. It is becoming clear that these interactions can be measured across many different omics layers, including the genome, transcriptome, epigenome, metabolome, and proteome, among others. Multi-omic studies of the microbiome can provide insight into the mechanisms underlying host-microbe interactions. As more omics layers are considered, increasingly sophisticated statistical methods are required to integrate them. In this review, we provide an overview of approaches currently used to characterize multi-omic interactions between host and microbiome data. While a large number of studies have generated a deeper understanding of host-microbiome interactions, there is still a need for standardization across approaches. Furthermore, microbiome studies would also benefit from the collection and curation of large, publicly available multi-omics datasets.

Statistical normalization methods in microbiome data with application to microbiome cancer research

Mounting evidence has shown that gut microbiome is associated with various cancers, including gastrointestinal (GI) tract and non-GI tract cancers. But microbiome data have unique characteristics and pose major challenges when using standard statistical methods causing results to be invalid or misleading. Thus, to analyze microbiome data, it not only needs appropriate statistical methods, but also requires microbiome data to be normalized prior to statistical analysis. Here, we first describe the unique characteristics of microbiome data and the challenges in analyzing them (Section 2). Then, we provide an overall review on the available normalization methods of 16S rRNA and shotgun metagenomic data along with examples of their applications in microbiome cancer research (Section 3). In Section 4, we comprehensively investigate how the normalization methods of 16S rRNA and shotgun metagenomic data are evaluated. Finally, we summarize and conclude with remarks on statistical normalization methods (Section 5). Altogether, this review aims to provide a broad and comprehensive view and remarks on the promises and challenges of the statistical normalization methods in microbiome data with microbiome cancer research examples.

Identification of cell-type-specific genes in multimodal single-cell data using deep neural network algorithm

The emergence of single-cell RNA sequencing (scRNA-seq) technology makes it possible to measure DNA, RNA, and protein in a single cell. Cellular Indexing of Transcriptomes and Epitopes by sequencing (CITE-seq) is a powerful multimodal single-cell research innovation, allowing researchers to capture RNA and surface protein expression on the same cells. Currently, identification of cell-type-specific genes in CITE-seq data is still challenging. In this study, we obtained a set of CITE-seq datasets from Kaggle database, which included the sequencing dataset of seven cell types during bone marrow stem cell differentiation. We used Student's t-test to analyze these transcription RNAs and pick out 133 significantly differentially expressed genes (DEGs) among all cell types. Functional enrichment revealed that these DEGs were strongly associated with blood-related diseases, providing important insights into the cellular heterogeneity within bone marrow stem cells. The relation between RNA and protein levels was performed by deep neural network (DNN) model and achieved a high prediction score of 0.867. Based on their coefficients in the DNN model, three genes (LGALS1, CENPV, TRIM24) were identified as cell-type-specific genes in erythrocyte progenitor. Our works provide a novel perspective regarding the differentiation of stem cells in the bone marrow and provide valuable insights for further research in this field.

Data pre-processing for analyzing microbiome data - A mini review

The human microbiome is an emerging research frontier due to its profound impacts on health. High-throughput microbiome sequencing enables studying microbial communities but suffers from analytical challenges. In particular, the lack of dedicated preprocessing methods to improve data quality impedes effective minimization of biases prior to downstream analysis. This review aims to address this gap by providing a comprehensive overview of preprocessing techniques relevant to microbiome research. We outline a typical workflow for microbiome data analysis. Preprocessing methods discussed include quality filtering, batch effect correction, imputation of missing values, normalization, and data transformation. We highlight strengths and limitations of each technique to serve as a practical guide for researchers and identify areas needing further methodological development. Establishing robust, standardized preprocessing will be essential for drawing valid biological conclusions from microbiome studies.

Critical review of 16S rRNA gene sequencing workflow in microbiome studies: From primer selection to advanced data analysis

The multi-batch reanalysis approach of jointly reevaluating gene/genome sequences from different works has gained particular relevance in the literature in recent years. The large amount of 16S ribosomal ribonucleic acid (rRNA) gene sequence data stored in public repositories and information in taxonomic databases of the same gene far exceeds that related to complete genomes. This review is intended to guide researchers new to studying microbiota, particularly the oral microbiota, using 16S rRNA gene sequencing and those who want to expand and update their knowledge to optimise their decision-making and improve their research results. First, we describe the advantages and disadvantages of using the 16S rRNA gene as a phylogenetic marker and the latest findings on the impact of primer pair selection on diversity and taxonomic assignment outcomes in oral microbiome studies. Strategies for primer selection based on these results are introduced. Second, we identified the key factors to consider in selecting the sequencing technology and platform. The process and particularities of the main steps for processing 16S rRNA gene-derived data are described in detail to enable researchers to choose the most appropriate bioinformatics pipeline and analysis methods based on the available evidence. We then produce an overview of the different types of advanced analyses, both the most widely used in the literature and the most recent approaches. Several indices, metrics and software for studying microbial communities are included, highlighting their advantages and disadvantages. Considering the principles of clinical metagenomics, we conclude that future research should focus on rigorous analytical approaches, such as developing predictive models to identify microbiome-based biomarkers to classify health and disease states. Finally, we address the batch effect concept and the microbiome-specific methods for accounting for or correcting them.

Machine learning approaches in microbiome research: challenges and best practices

Microbiome data predictive analysis within a machine learning (ML) workflow presents numerous domain-specific challenges involving preprocessing, feature selection, predictive modeling, performance estimation, model interpretation, and the extraction of biological information from the results. To assist decision-making, we offer a set of recommendations on algorithm selection, pipeline creation and evaluation, stemming from the COST Action ML4Microbiome. We compared the suggested approaches on a multi-cohort shotgun metagenomics dataset of colorectal cancer patients, focusing on their performance in disease diagnosis and biomarker discovery. It is demonstrated that the use of compositional transformations and filtering methods as part of data preprocessing does not always improve the predictive performance of a model. In contrast, the multivariate feature selection, such as the Statistically Equivalent Signatures algorithm, was effective in reducing the classification error. When validated on a separate test dataset, this algorithm in combination with random forest modeling, provided the most accurate performance estimates. Lastly, we showed how linear modeling by logistic regression coupled with visualization techniques such as Individual Conditional Expectation (ICE) plots can yield interpretable results and offer biological insights. These findings are significant for clinicians and non-experts alike in translational applications.

A prebiotic diet modulates the oral microbiome composition and results in the attenuation of oropharyngeal candidiasis in mice

Oral bacteria can influence the ability of Candida albicans to cause oropharyngeal candidiasis (OPC). We recently reported that a Lactobacillus johnsonii-enriched oral microbiota reduced C. albicans virulence in an immunosuppressed OPC mouse model. As a follow-up, in this work, we aimed to enrich the resident oral Lactobacillus communities with a prebiotic diet to further assess their effect on the severity of OPC. We tested the effect of a prebiotic xylo-oligosaccharides (XOS)-enriched diet in the oral global bacterial composition and severity of OPC. We assessed changes in the oral microbiome composition via 16S-rRNA gene high-throughput sequencing, validated by qPCR. The impact of the prebiotic diet on Candida infection was assessed by quantifying changes in oral fungal and bacterial biomass and scoring tongue lesions. Contrary to expectations, oral Lactobacillus communities were not enriched by the XOS-supplemented diet. Yet, XOS modulated the oral microbiome composition, increasing Bifidobacterium abundance and reducing enterococci and staphylococci. In the OPC model, the XOS diet attenuated Candida virulence and bacterial dysbiosis, increasing lactobacilli and reducing enterococci on the oral mucosa. We conclude that XOS attenuates Candida virulence by promoting a bacterial microbiome structure more resilient to Candida infection. IMPORTANCE This is the first study on the effects of a prebiotic diet on the oral mucosal bacterial microbiome and an oropharyngeal candidiasis (OPC) mouse model. We found that xylo-oligosaccharides change the oral bacterial community composition and attenuate OPC. Our results contribute to the understanding of the impact of the oral bacterial communities on Candida virulence.

How missing value imputation is confounded with batch effects and what you can do about it

In data-processing pipelines, upstream steps can influence downstream processes because of their sequential nature. Among these data-processing steps, batch effect (BE) correction (BEC) and missing value imputation (MVI) are crucial for ensuring data suitability for advanced modeling and reducing the likelihood of false discoveries. Although BEC-MVI interactions are not well studied, they are ultimately interdependent. Batch sensitization can improve the quality of MVI. Conversely, accounting for missingness also improves proper BE estimation in BEC. Here, we discuss how BEC and MVI are interconnected and interdependent. We show how batch sensitization can improve any MVI and bring attention to the idea of BE-associated missing values (BEAMs). Finally, we discuss how batch-class imbalance problems can be mitigated by borrowing ideas from machine learning.

Multi-level analysis of the gut-brain axis shows autism spectrum disorder-associated molecular and microbial profiles

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by heterogeneous cognitive, behavioral and communication impairments. Disruption of the gut-brain axis (GBA) has been implicated in ASD although with limited reproducibility across studies. In this study, we developed a Bayesian differential ranking algorithm to identify ASD-associated molecular and taxa profiles across 10 cross-sectional microbiome datasets and 15 other datasets, including dietary patterns, metabolomics, cytokine profiles and human brain gene expression profiles. We found a functional architecture along the GBA that correlates with heterogeneity of ASD phenotypes, and it is characterized by ASD-associated amino acid, carbohydrate and lipid profiles predominantly encoded by microbial species in the genera Prevotella, Bifidobacterium, Desulfovibrio and Bacteroides and correlates with brain gene expression changes, restrictive dietary patterns and pro-inflammatory cytokine profiles. The functional architecture revealed in age-matched and sex-matched cohorts is not present in sibling-matched cohorts. We also show a strong association between temporal changes in microbiome composition and ASD phenotypes. In summary, we propose a framework to leverage multi-omic datasets from well-defined cohorts and investigate how the GBA influences ASD.

Statistical analysis of high-dimensional biomedical data: a gentle introduction to analytical goals, common approaches and challenges

BACKGROUND

In high-dimensional data (HDD) settings, the number of variables associated with each observation is very large. Prominent examples of HDD in biomedical research include omics data with a large number of variables such as many measurements across the genome, proteome, or metabolome, as well as electronic health records data that have large numbers of variables recorded for each patient. The statistical analysis of such data requires knowledge and experience, sometimes of complex methods adapted to the respective research questions.

METHODS

Advances in statistical methodology and machine learning methods offer new opportunities for innovative analyses of HDD, but at the same time require a deeper understanding of some fundamental statistical concepts. Topic group TG9 "High-dimensional data" of the STRATOS (STRengthening Analytical Thinking for Observational Studies) initiative provides guidance for the analysis of observational studies, addressing particular statistical challenges and opportunities for the analysis of studies involving HDD. In this overview, we discuss key aspects of HDD analysis to provide a gentle introduction for non-statisticians and for classically trained statisticians with little experience specific to HDD.

RESULTS

The paper is organized with respect to subtopics that are most relevant for the analysis of HDD, in particular initial data analysis, exploratory data analysis, multiple testing, and prediction. For each subtopic, main analytical goals in HDD settings are outlined. For each of these goals, basic explanations for some commonly used analysis methods are provided. Situations are identified where traditional statistical methods cannot, or should not, be used in the HDD setting, or where adequate analytic tools are still lacking. Many key references are provided.

CONCLUSIONS

This review aims to provide a solid statistical foundation for researchers, including statisticians and non-statisticians, who are new to research with HDD or simply want to better evaluate and understand the results of HDD analyses.

Effects of age, seasonality, and reproductive status on the gut microbiome of Southern White Rhinoceros (Ceratotherium simum simum) at the North Carolina zoo

BACKGROUND

Managed southern white rhinoceros (Ceratotherium simum simum) serve as assurance populations for wild conspecifics threatened by poaching and other anthropocentric effects, though many managed populations experience subfertility and reproductive failure. Gut microbiome and host health are inextricably linked, and reproductive outcomes in managed southern white rhinoceros may be mediated in part by their diet and gut microbial diversity. Thus, understanding microbial dynamics within managed populations may help improve conservation efforts. We characterized the taxonomic composition of the gut microbiome in the managed population of female southern white rhinoceros (n = 8) at the North Carolina Zoo and investigated the effects of seasonality (summer vs. winter) and age classes (juveniles (n = 2; 0-2 years), subadults (n = 2; 3-7 years), and adults (n = 4; >7 years)) on microbial richness and community structure. Collection of a fecal sample was attempted for each individual once per month from July-September 2020 and January-March 2021 resulting in a total of 41 samples analyzed. Microbial DNA was extracted and sequenced using the V3-V4 region of the 16S rRNA bacterial gene. Total operational taxonomic units (OTUs), alpha diversity (species richness, Shannon diversity), and beta diversity (Bray-Curtis dissimilarity, linear discriminant analysis effect size) indices were examined, and differentially enriched taxa were identified.

RESULTS

There were differences (p < 0.05) in alpha and beta diversity indices across individuals, age groups, and sampling months. Subadult females had higher levels of Shannon diversity (Wilcoxon, p < 0.05) compared to adult females and harbored a community cluster distinct from both juveniles and adults. Samples collected during winter months (January-March 2021) possessed higher species richness and statistically distinct communities compared to summer months (July-September 2020) (PERMANOVA, p < 0.05). Reproductively active (n = 2) and currently nonreproductive adult females (n = 2) harbored differentially enriched taxa, with the gut microbiome of nonreproductive females significantly enriched (p = 0.001) in unclassified members of Mobiluncus, a genus which possesses species associated with poor reproductive outcomes in other animal species when identified in the cervicovaginal microbiome.

CONCLUSION

Together, our results increase the understanding of age and season related microbial variation in southern white rhinoceros at the North Carolina Zoo and have identified a potential microbial biomarker for reproductive concern within managed female southern white rhinoceros.

MBECS: Microbiome Batch Effects Correction Suite

Despite the availability of batch effect correcting algorithms (BECA), no comprehensive tool that combines batch correction and evaluation of the results exists for microbiome datasets. This work outlines the Microbiome Batch Effects Correction Suite development that integrates several BECAs and evaluation metrics into a software package for the statistical computation framework R.

Concordance-Based Batch Effect Correction for Large-Scale Metabolomics

For a large-scale metabolomics study, sample collection, preparation, and analysis may last several days, months, or even (intermittently) over years. This may lead to apparent batch effects in the acquired metabolomics data due to variability in instrument status, environmental conditions, or experimental operators. Batch effects may confound the true biological relationships among metabolites and thus obscure real metabolic changes. At present, most of the commonly used batch effect correction (BEC) methods are based on quality control (QC) samples, which require sufficient and stable QC samples. However, the quality of the QC samples may deteriorate if the experiment lasts for a long time. Alternatively, isotope-labeled internal standards have been used, but they generally do not provide good coverage of the metabolome. On the other hand, BEC can also be conducted through a data-driven method, in which no QC sample is needed. Here, we propose a novel data-driven BEC method, namely, CordBat, to achieve concordance between each batch of samples. In the proposed CordBat method, a reference batch is first selected from all batches of data, and the remaining batches are referred to as "other batches." The reference batch serves as the baseline for the batch adjustment by providing a coordinate of correlation between metabolites. Next, a Gaussian graphical model is built on the combined dataset of reference and other batches, and finally, BEC is achieved by optimizing the correction coefficients in the other batches so that the correlation between metabolites of each batch and their combinations are in concordance with that of the reference batch. Three real-world metabolomics datasets are used to evaluate the performance of CordBat by comparing it with five commonly used BEC methods. The present experimental results showed the effectiveness of CordBat in batch effect removal and the concordance of correlation between metabolites after BEC. CordBat was found to be comparable to the QC-based methods and achieved better performance in the preservation of biological effects. The proposed CordBat method may serve as an alternative BEC method for large-scale metabolomics that lack proper QC samples.

PLSDA-batch: a multivariate framework to correct for batch effects in microbiome data

Microbial communities are highly dynamic and sensitive to changes in the environment. Thus, microbiome data are highly susceptible to batch effects, defined as sources of unwanted variation that are not related to and obscure any factors of interest. Existing batch effect correction methods have been primarily developed for gene expression data. As such, they do not consider the inherent characteristics of microbiome data, including zero inflation, overdispersion and correlation between variables. We introduce new multivariate and non-parametric batch effect correction methods based on Partial Least Squares Discriminant Analysis (PLSDA). PLSDA-batch first estimates treatment and batch variation with latent components, then subtracts batch-associated components from the data. The resulting batch-effect-corrected data can then be input in any downstream statistical analysis. Two variants are proposed to handle unbalanced batch x treatment designs and to avoid overfitting when estimating the components via variable selection. We compare our approaches with popular methods managing batch effects, namely, removeBatchEffect, ComBat and Surrogate Variable Analysis, in simulated and three case studies using various visual and numerical assessments. We show that our three methods lead to competitive performance in removing batch variation while preserving treatment variation, especially for unbalanced batch $\times $ treatment designs. Our downstream analyses show selections of biologically relevant taxa. This work demonstrates that batch effect correction methods can improve microbiome research outputs. Reproducible code and vignettes are available on GitHub.

BIRDMAn: A Bayesian differential abundance framework that enables robust inference of host-microbe associations

Quantifying the differential abundance (DA) of specific taxa among experimental groups in microbiome studies is challenging due to data characteristics (e.g., compositionality, sparsity) and specific study designs (e.g., repeated measures, meta-analysis, cross-over). Here we present BIRDMAn (Bayesian Inferential Regression for Differential Microbiome Analysis), a flexible DA method that can account for microbiome data characteristics and diverse experimental designs. Simulations show that BIRDMAn models are robust to uneven sequencing depth and provide a >20-fold improvement in statistical power over existing methods. We then use BIRDMAn to identify antibiotic-mediated perturbations undetected by other DA methods due to subject-level heterogeneity. Finally, we demonstrate how BIRDMAn can construct state-of-the-art cancer-type classifiers using The Cancer Genome Atlas (TCGA) dataset, with substantial accuracy improvements over random forests and existing DA tools across multiple sequencing centers. Collectively, BIRDMAn extracts more informative biological signals while accounting for study-specific experimental conditions than existing approaches.

Temporal shifts of the microbiome associated with antibiotic treatment of purpuric drug eruptions related to epidermal growth factor receptor inhibitors

BACKGROUND

Epidermal growth factor receptor (EGFR) inhibitors are selective and effective treatments for cancers with relevant mutations. Purpuric drug eruptions are an uncommon but clinically significant dermatological side effect related to EGFR inhibitor use that are associated with positive bacterial cultures and responsive to antibiotic treatment. However, the longitudinal temporal shifts in the skin microbiome that occur before and after antibiotic treatment of purpuric drug eruptions remain largely unknown.

OBJECTIVES

To characterize temporal changes in the skin and mucosal microbiomes before and after antibiotic treatment of EGFR inhibitor-related purpuric drug eruptions.

METHODS

Twelve patients who experienced EGFR inhibitor-related purpuric drug eruptions were recruited from a dermato-oncology clinic in Taiwan from May 2017 to April 2018. Swabs were obtained from skin lesions and the nasal mucosa before and after antibiotic treatment of purpuric drug eruptions. After the amplification and sequencing of bacterial 16S rRNA genes, the diversity and compositions of microbiomes sampled at different time points were compared.

RESULTS

The alpha diversity (represented by the Shannon index) of the skin microbiome increased significantly in the recovered phase of purpuric drug eruptions compared with that of the active phase. By contrast, the nasal microbiome showed no significant change in alpha diversity. The relative abundance of Staphylococcus significantly decreased in samples from skin of the recovered phase, which was confirmed by analysis of compositions of microbiomes (ANCOM) and the ALDEx2 analysis packages in R.

CONCLUSIONS

The cutaneous microbiome of purpuric drug eruptions showed a significant increase in alpha diversity and a decrease in the relative abundance of Staphylococcus following antibiotic treatment. These findings may help guide antimicrobial therapy of this EGFR inhibitor-related condition.

Compositionality, sparsity, spurious heterogeneity, and other data-driven challenges for machine learning algorithms within plant microbiome studies

The plant-associated microbiome is a key component of plant systems, contributing to their health, growth, and productivity. The application of machine learning (ML) in this field promises to help untangle the relationships involved. However, measurements of microbial communities by high-throughput sequencing pose challenges for ML. Noise from low sample sizes, soil heterogeneity, and technical factors can impact the performance of ML. Additionally, the compositional and sparse nature of these datasets can impact the predictive accuracy of ML. We review recent literature from plant studies to illustrate that these properties often go unmentioned. We expand our analysis to other fields to quantify the degree to which mitigation approaches improve the performance of ML and describe the mathematical basis for this. With the advent of accessible analytical packages for microbiome data including learning models, researchers must be familiar with the nature of their datasets.

The structural discrepancy between the small and large gut microbiota of Asiatic toad (Bufo gargarizans) during hibernation

Hibernating amphibians are suitable for the research on the adaptation of gut microbiota to long-term fasting and cold stresses. However, the previous studies mainly focus on the large or whole gut microbiota but not the small gut microbiota. To test the structural discrepancy between the small and large gut microbiota during hibernation, we performed two independent batches of 16S rRNA gene amplicon sequencing to profile the small and large gut microbiota of hibernating Asiatic toad (Bufo gargarizans) from two wild populations. Both batches of data revealed that Proteobacteria, Bacteroidetes, and Firmicutes were the three most dominant phyla in the small and large gut microbiota. Three core OTUs with 100% occurrence in all gut microbiotas were annotated as Pseudomonas. A significant structural discrepancy was detected between the small and large gut microbiota. For instance, Proteobacteria assembled in the small intestine with a higher proportion than it did in the large intestine, but Bacteroidetes and Firmicutes assembled in the large intestine with a higher proportion than they did in the small intestine. The large gut microbiota exhibited higher diversity than the small gut microbiota. Nevertheless, a severe batch effect existed in the structural analysis of the gut microbiotas. The large gut microbiota showed a better resistance to the batch effect than the small gut microbiota did. This study provides preliminary evidence that microbes assemble in the small and large intestines of amphibians with discrepant patterns during hibernation.

Capacity of deep-sea corals to obtain nutrition from cold seeps aligned with microbiome reorganization

Cold seeps in the deep sea harbor various animals that have adapted to utilize seepage chemicals with the aid of chemosynthetic microbes that serve as primary producers. Corals are among the animals that live near seep habitats and yet, there is a lack of evidence that corals gain benefits and/or incur costs from cold seeps. Here, we focused on Callogorgia delta and Paramuricea sp. type B3 that live near and far from visual signs of currently active seepage at five sites in the deep Gulf of Mexico. We tested whether these corals rely on chemosynthetically-derived food in seep habitats and how the proximity to cold seeps may influence; (i) coral colony traits (i.e., health status, growth rate, regrowth after sampling, and branch loss) and associated epifauna, (ii) associated microbiome, and (iii) host transcriptomes. Stable isotope data showed that many coral colonies utilized chemosynthetically derived food, but the feeding strategy differed by coral species. The microbiome composition of C. delta, unlike Paramuricea sp., varied significantly between seep and non-seep colonies and both coral species were associated with various sulfur-oxidizing bacteria (SUP05). Interestingly, the relative abundances of SUP05 varied among seep and non-seep colonies and were strongly correlated with carbon and nitrogen stable isotope values. In contrast, the proximity to cold seeps did not have a measurable effect on gene expression, colony traits, or associated epifauna in coral species. Our work provides the first evidence that some corals may gain benefits from living near cold seeps with apparently limited costs to the colonies. Cold seeps provide not only hard substrate but also food to cold-water corals. Furthermore, restructuring of the microbiome communities (particularly SUP05) is likely the key adaptive process to aid corals in utilizing seepage-derived carbon. This highlights that those deep-sea corals may upregulate particular microbial symbiont communities to cope with environmental gradients.

Accumulation of Black Carbon Particles in Placenta, Cord Blood, and Childhood Urine in Association with the Intestinal Microbiome Diversity and Composition in Four- to Six-Year-Old Children in the ENVIRONAGE Birth Cohort

BACKGROUND

The gut microbiome plays an essential role in human health. Despite the link between air pollution exposure and various diseases, its association with the gut microbiome during susceptible life periods remains scarce.

OBJECTIVES

In this study, we examined the association between black carbon particles quantified in prenatal and postnatal biological matrices and bacterial richness and diversity measures, and bacterial families.

METHODS

A total of 85 stool samples were collected from 4- to 6-y-old children enrolled in the ENVIRonmental influence ON early AGEing birth cohort. We performed 16S rRNA gene sequencing to calculate bacterial richness and diversity indices (Chao1 richness, Shannon diversity, and Simpson diversity) and the relative abundance of bacterial families. Black carbon particles were quantified via white light generation under femtosecond pulsed laser illumination in placental tissue and cord blood, employed as prenatal exposure biomarkers, and in urine, used as a post-natal exposure biomarker. We used robust multivariable-adjusted linear models to examine the associations between quantified black carbon loads and measures of richness (Chao1 index) and diversity (Shannon and Simpson indices), adjusting for parity, season of delivery, sequencing batch, age, sex, weight and height of the child, and maternal education. Additionally, we performed a differential relative abundance analysis of bacterial families with a correction for sampling fraction bias. Results are expressed as percentage difference for a doubling in black carbon loads with 95% confidence interval (CI).

RESULTS

Two diversity indices were negatively associated with placental black carbon [Shannon: -4.38% (95% CI: -8.31%, -0.28%); Simpson: -0.90% (95% CI: -1.76%, -0.04%)], cord blood black carbon [Shannon: -3.38% (95% CI: -5.66%, -0.84%); Simpson: -0.91 (95% CI: -1.66%, -0.16%)], and urinary black carbon [Shannon: -3.39% (95% CI: -5.77%, -0.94%); Simpson: -0.89% (95% CI: -1.37%, -0.40%)]. The explained variance of black carbon on the above indices varied from 6.1% to 16.6%. No statistically significant associations were found between black carbon load and the Chao1 richness index. After multiple testing correction, placental black carbon was negatively associated with relative abundance of the bacterial families Defluviitaleaceae and Marinifilaceae, and urinary black carbon with Christensenellaceae and Coriobacteriaceae; associations with cord blood black carbon were not statistically significant after correction.

CONCLUSION

Black carbon particles quantified in prenatal and postnatal biological matrices were associated with the composition and diversity of the childhood intestinal microbiome. These findings address the influential role of exposure to air pollution during pregnancy and early life in human health. https://doi.org/10.1289/EHP11257.

The gut microbiome in human health and disease-Where are we and where are we going? A bibliometric analysis

Background

There are trillions of microbiota in our intestinal tract, and they play a significant role in health and disease via interacting with the host in metabolic, immune, neural, and endocrine pathways. Over the past decades, numerous studies have been published in the field of gut microbiome and disease. Although there are narrative reviews of gut microbiome and certain diseases, the whole field is lack of systematic and quantitative analysis. Therefore, we outline research status of the gut microbiome and disease, and present insights into developments and characteristics of this field to provide a holistic grasp and future research directions.

Methods

An advanced search was carried out in the Web of Science Core Collection (WoSCC), basing on the term "gut microbiome" and its synonyms. The current status and developing trends of this scientific domain were evaluated by bibliometric methodology. CiteSpace was used to perform collaboration network analysis, co-citation analysis and citation burst detection.

Results

A total of 29,870 articles and 13,311 reviews were retrieved from the database, which involve 42,900 keywords, 176 countries/regions, 19,065 institutions, 147,225 authors and 4,251 journals. The gut microbiome and disease research is active and has received increasing attention. Co-cited reference analysis revealed the landmark articles in the field. The United States had the largest number of publications and close cooperation with other countries. The current research mainly focuses on gastrointestinal diseases, such as inflammatory bowel disease (IBD), ulcerative colitis (UC) and Crohn's disease (CD), while extra-intestinal diseases are also rising, such as obesity, diabetes, cardiovascular disease, Alzheimer's disease, Parkinson's disease. Omics technologies, fecal microbiota transplantation (FMT) and metabolites linked to mechanism would be more concerned in the future.

Conclusion

The gut microbiome and disease has been a booming field of research, and the trend is expected to continue. Overall, this research field shows a multitude of challenges and great opportunities.

Multiomic analysis reveals microbiome-related relationships between cocaine use and metabolites

OBJECTIVE

Over 19 million individuals globally have a cocaine use disorder, a significant public health crisis. Cocaine has also been associated with a pro-inflammatory state and recently with imbalances in the intestinal microbiota as compared to nonuse. The objective of this pilot study was to characterize the gut microbiota and plasma metabolites in people with HIV (PWH) who use cocaine compared with those who do not.

DESIGN

Cross-sectional study.

METHODS

A pilot study in PWH was conducted on 25 cocaine users and 25 cocaine nonusers from the Miami Adult Studies on HIV cohort. Stool samples and blood plasma were collected. Bacterial composition was characterized using 16S rRNA sequencing. Metabolomics in plasma were determined using gas and liquid chromatography/mass spectrometry.

RESULTS

The relative abundances of the Lachnopspira genus, Oscillospira genus, Bifidobacterium adolescentis species, and Euryarchaeota phylum were significantly higher in the cocaine- using PWH compared to cocaine-nonusing PWH. Cocaine-use was associated with higher levels of several metabolites: products of dopamine catabolism (3-methoxytyrosine and 3-methoxytyramine sulfate), phenylacetate, benzoate, butyrate, and butyrylglycine.

CONCLUSIONS

Cocaine use was associated with higher abundances of taxa and metabolites known to be associated with pathogenic states that include gastrointestinal conditions. Understanding key intestinal bacterial functional pathways that are altered due to cocaine use in PWH will provide a better understanding of the relationships between the host intestinal microbiome and potentially provide novel treatments to improve health.

Gut Microbiota in Nutrition and Health with a Special Focus on Specific Bacterial Clusters

Health is influenced by how the gut microbiome develops as a result of external and internal factors, such as nutrition, the environment, medication use, age, sex, and genetics. Alpha and beta diversity metrics and (enterotype) clustering methods are commonly employed to perform population studies and to analyse the effects of various treatments, yet, with the continuous development of (new) sequencing technologies, and as various omics fields as a result become more accessible for investigation, increasingly sophisticated methodologies are needed and indeed being developed in order to disentangle the complex ways in which the gut microbiome and health are intertwined. Diseases of affluence, such as type 2 diabetes (T2D) and cardiovascular diseases (CVD), are commonly linked to species associated with the Bacteroides enterotype(s) and a decline of various (beneficial) complex microbial trophic networks, which are in turn linked to the aforementioned factors. In this review, we (1) explore the effects that some of the most common internal and external factors have on the gut microbiome composition and how these in turn relate to T2D and CVD, and (2) discuss research opportunities enabled by and the limitations of some of the latest technical developments in the microbiome sector, including the use of artificial intelligence (AI), strain tracking, and peak to trough ratios.

Batch effects removal for microbiome data via conditional quantile regression

Batch effects in microbiome data arise from differential processing of specimens and can lead to spurious findings and obscure true signals. Strategies designed for genomic data to mitigate batch effects usually fail to address the zero-inflated and over-dispersed microbiome data. Most strategies tailored for microbiome data are restricted to association testing or specialized study designs, failing to allow other analytic goals or general designs. Here, we develop the Conditional Quantile Regression (ConQuR) approach to remove microbiome batch effects using a two-part quantile regression model. ConQuR is a comprehensive method that accommodates the complex distributions of microbial read counts by non-parametric modeling, and it generates batch-removed zero-inflated read counts that can be used in and benefit usual subsequent analyses. We apply ConQuR to simulated and real microbiome datasets and demonstrate its advantages in removing batch effects while preserving the signals of interest.

Characterization of the oral and fecal microbiota associated with atopic dermatitis in dogs selected from a purebred Shiba Inu colony

Atopic dermatitis (AD) is a chronic and relapsing multifactorial inflammatory skin disease that also affects dogs. The oral and gut microbiota are associated with many disorders, including allergy. Few studies have addressed the oral and gut microbiota in dogs, although the skin microbiota has been studied relatively well in these animals. Here, we studied the AD-associated oral and gut microbiota in 16 healthy and nine AD dogs from a purebred Shiba Inu colony. We found that the diversity of the oral microbiota was significantly different among the dogs, whereas no significant difference was observed in the gut microbiota. Moreover, a differential abundance analysis detected the Family_XIII_AD3011_group (Anaerovoracaceae) in the gut microbiota of AD dogs; however, no bacterial taxa were detected in the oral microbiota. Third, the comparison of the microbial co-occurrence patterns between AD and healthy dogs identified differential networks in which the bacteria in the oral microbiota that were most strongly associated with AD were related with human periodontitis, whereas those in the gut microbiota were related with dysbiosis and gut inflammation. These results suggest that AD can alter the oral and gut microbiota in dogs.

Prediction of Smoking Habits From Class-Imbalanced Saliva Microbiome Data Using Data Augmentation and Machine Learning

Human microbiome research is moving from characterization and association studies to translational applications in medical research, clinical diagnostics, and others. One of these applications is the prediction of human traits, where machine learning (ML) methods are often employed, but face practical challenges. Class imbalance in available microbiome data is one of the major problems, which, if unaccounted for, leads to spurious prediction accuracies and limits the classifier's generalization. Here, we investigated the predictability of smoking habits from class-imbalanced saliva microbiome data by combining data augmentation techniques to account for class imbalance with ML methods for prediction. We collected publicly available saliva 16S rRNA gene sequencing data and smoking habit metadata demonstrating a serious class imbalance problem, i.e., 175 current vs. 1,070 non-current smokers. Three data augmentation techniques (synthetic minority over-sampling technique, adaptive synthetic, and tree-based associative data augmentation) were applied together with seven ML methods: logistic regression, k-nearest neighbors, support vector machine with linear and radial kernels, decision trees, random forest, and extreme gradient boosting. K-fold nested cross-validation was used with the different augmented data types and baseline non-augmented data to validate the prediction outcome. Combining data augmentation with ML generally outperformed baseline methods in our dataset. The final prediction model combined tree-based associative data augmentation and support vector machine with linear kernel, and achieved a classification performance expressed as Matthews correlation coefficient of 0.36 and AUC of 0.81. Our method successfully addresses the problem of class imbalance in microbiome data for reliable prediction of smoking habits.

Statistical challenges in longitudinal microbiome data analysis

The microbiome is a complex and dynamic community of microorganisms that co-exist interdependently within an ecosystem, and interact with its host or environment. Longitudinal studies can capture temporal variation within the microbiome to gain mechanistic insights into microbial systems; however, current statistical methods are limited due to the complex and inherent features of the data. We have identified three analytical objectives in longitudinal microbial studies: (1) differential abundance over time and between sample groups, demographic factors or clinical variables of interest; (2) clustering of microorganisms evolving concomitantly across time and (3) network modelling to identify temporal relationships between microorganisms. This review explores the strengths and limitations of current methods to fulfill these objectives, compares different methods in simulation and case studies for objectives (1) and (2), and highlights opportunities for further methodological developments. R tutorials are provided to reproduce the analyses conducted in this review.

Variation in Rumen Bacteria of Lacaune Dairy Ewes From One Week to the Next

Bacteria are the most abundant microorganisms in the rumen microbiota and play essential roles, mainly fermenting plant compounds that yield fatty acids. In this study, we aimed at assessing stability of both bacterial composition and of its associations with rumen and milk fatty acids phenotypes over a 1-week period. The study was performed using 118 Lacaune dairy ewes from the INRAE Experimental Unit of La Fage. Rumen and milk samples were obtained from the ewes twice, 1 week apart, and microbiota composition, volatile and long-chain fatty acid concentrations were analyzed. Bacterial composition was assessed using 16S rRNA gene sequencing, and microbiota and fatty acids were analyzed as compositional data. As we worked with relative abundances expressed in a constrained space, the centered log-ratio transformation enabled to transform data to work with multivariate analyses in the Euclidian space. Bacterial composition differed between the 2 weeks of sampling, characterized by different proportions of the two main phyla, Bacteroidetes and Firmicutes. The repeatability of the operational taxonomic units (OTUs) was low, although it varied significantly. However, 66 of them presented a repeatability of over 0.50 and were particularly associated with fatty acid phenotypes. Even though the OTUs from the same bacterial families presented similar correlations to fatty acids in both weeks, only a few OTUs were conserved over the 2 weeks. We proved with the help of sequencing data that there is significant change in microbial composition over a week in terms of abundance of different families of bacteria. Further studies are required to determine the impact of bacterial composition alterations over 1 week, and the specificities of the highly repeatable OTUs.

Gut Microbiome Signatures in the Progression of Hepatitis B Virus-Induced Liver Disease

The gut microbiome is associated with hepatitis B virus (HBV)-induced liver disease, which progresses from chronic hepatitis B, to liver cirrhosis, and eventually to hepatocellular carcinoma. Studies have analyzed the gut microbiome at each stage of HBV-induced liver diseases, but a consensus has not been reached on the microbial signatures across these stages. Here, we conducted by a systematic meta-analysis of 486 fecal samples from publicly available 16S rRNA gene datasets across all disease stages, and validated the results by a gut microbiome characterization on an independent cohort of 15 controls, 23 chronic hepatitis B, 20 liver cirrhosis, and 22 hepatocellular carcinoma patients. The integrative analyses revealed 13 genera consistently altered at each of the disease stages both in public and validation datasets, suggesting highly robust microbiome signatures. Specifically, Colidextribacter and Monoglobus were enriched in healthy controls. An unclassified Lachnospiraceae genus was specifically elevated in chronic hepatitis B, whereas Bilophia was depleted. Prevotella and Oscillibacter were depleted in liver cirrhosis. And Coprococcus and Faecalibacterium were depleted in hepatocellular carcinoma. Classifiers established using these 13 genera showed diagnostic power across all disease stages in a cross-validation between public and validation datasets (AUC = 0.65–0.832). The identified microbial taxonomy serves as non-invasive biomarkers for monitoring the progression of HBV-induced liver disease, and may contribute to microbiome-based therapies.

Identification of shared and disease-specific host gene-microbiome associations across human diseases using multi-omic integration

While gut microbiome and host gene regulation independently contribute to gastrointestinal disorders, it is unclear how the two may interact to influence host pathophysiology. Here we developed a machine learning-based framework to jointly analyse paired host transcriptomic (n = 208) and gut microbiome (n = 208) profiles from colonic mucosal samples of patients with colorectal cancer, inflammatory bowel disease and irritable bowel syndrome. We identified associations between gut microbes and host genes that depict shared as well as disease-specific patterns. We found that a common set of host genes and pathways implicated in gastrointestinal inflammation, gut barrier protection and energy metabolism are associated with disease-specific gut microbes. Additionally, we also found that mucosal gut microbes that have been implicated in all three diseases, such as Streptococcus, are associated with different host pathways in each disease, suggesting that similar microbes can affect host pathophysiology in a disease-specific manner through regulation of different host genes. Our framework can be applied to other diseases for the identification of host gene-microbiome associations that may influence disease outcomes.

Large-scale microbiome data integration enables robust biomarker identification

The close association between gut microbiota dysbiosis and human diseases is being increasingly recognized. However, contradictory results are frequently reported, as confounding effects exist. The lack of unbiased data integration methods is also impeding the discovery of disease-associated microbial biomarkers from different cohorts. Here we propose an algorithm, NetMoss, for assessing shifts of microbial network modules to identify robust biomarkers associated with various diseases. Compared to previous approaches, the NetMoss method shows better performance in removing batch effects. Through comprehensive evaluations on both simulated and real datasets, we demonstrate that NetMoss has great advantages in the identification of disease-related biomarkers. Based on analysis of pandisease microbiota studies, there is a high prevalence of multidisease-related bacteria in global populations. We believe that large-scale data integration will help in understanding the role of the microbiome from a more comprehensive perspective and that accurate biomarker identification will greatly promote microbiome-based medical diagnosis.

Learning models for colorectal cancer signature reconstruction and classification in patients with chronic inflammatory bowel disease

BACKGROUND

In their everyday life, clinicians face an overabundance of biological indicators potentially helpful during a disease therapy. In this context, to be able to reliably identify a reduced number of those markers showing the ability of optimising the classification of treatment outcomes becomes a factor of vital importance to medical prognosis. In this work, we focus our interest in inflammatory bowel disease (IBD), a long-life threaten with a continuous increasing prevalence worldwide. In particular, IBD can be described as a set of autoimmune conditions affecting the gastrointestinal tract whose two main types are Crohn’s disease and ulcerative colitis.

AIM

To identify the minimal signature of microRNA (miRNA) associated with colorectal cancer (CRC) in patients with one chronic IBD.

METHODS

We provide a framework of well-established statistical and computational learning methods wisely adapted to reconstructing a CRC network leveraged to stratify these patients.

RESULTS

Our strategy resulted in an adjusted signature of 5 miRNAs out of approximately 2600 in Crohn’s Disease (resp. 8 in Ulcerative Colitis) with a percentage of success in patient classification of 82% (resp. 81%).

CONCLUSION

Importantly, these two signatures optimally balance the proportion between the number of significant miRNAs and their percentage of success in patients’ stratification.

Ontology-aware neural network: a general framework for pattern mining from microbiome data

With the rapid accumulation of microbiome data around the world, numerous computational bioinformatics methods have been developed for pattern mining from such paramount microbiome data. Current microbiome data mining methods, such as gene and species mining, rely heavily on sequence comparison. Most of these methods, however, have a clear trade-off, particularly, when it comes to big-data analytical efficiency and accuracy. Microbiome entities are usually organized in ontology structures, and pattern mining methods that have considered ontology structures could offer advantages in mining efficiency and accuracy. Here, we have summarized the ontology-aware neural network (ONN) as a novel framework for microbiome data mining. We have discussed the applications of ONN in multiple contexts, including gene mining, species mining and microbial community dynamic pattern mining. We have then highlighted one of the most important characteristics of ONN, namely, novel knowledge discovery, which makes ONN a standout among all microbiome data mining methods. Finally, we have provided several applications to showcase the advantage of ONN over other methods in microbiome data mining. In summary, ONN represents a paradigm shift for pattern mining from microbiome data: from traditional machine learning approach to ontology-aware and model-based approach, which has found its broad application scenarios in microbiome data mining.

What is unknown in using microbiota as a therapeutic?

Fecal microbiota transplantation (FMT) has been used extensively in the treatment of various gastrointestinal and extraintestinal conditions, despite that there are still a lot of missing gaps in our knowledge in the gut microbiota and its behavior. This article describes the unknowns in microbiota biology (undetected microbes, uncertain colonization, unclear mechanisms of action, uncertain indications, unsure long-term efficacy, or side effects). We discuss how these unknowns may affect the therapeutic uses of FMT, and the potentials and caveats of other related microbiota-based therapies. When used as an experimental therapy or last resort in difficult conditions, caution should be taken against inadvertent complications. Clear documentations of post-treatment events should be made mandatory, classified, and graded as in clinical trials. Further robust scientific experiments and properly designed clinical studies are needed.

Benchmark of Data Processing Methods and Machine Learning Models for Gut Microbiome-Based Diagnosis of Inflammatory Bowel Disease

Patients with inflammatory bowel disease (IBD) wait months and undergo numerous invasive procedures between the initial appearance of symptoms and receiving a diagnosis. In order to reduce time until diagnosis and improve patient wellbeing, machine learning algorithms capable of diagnosing IBD from the gut microbiome's composition are currently being explored. To date, these models have had limited clinical application due to decreased performance when applied to a new cohort of patient samples. Various methods have been developed to analyze microbiome data which may improve the generalizability of machine learning IBD diagnostic tests. With an abundance of methods, there is a need to benchmark the performance and generalizability of various machine learning pipelines (from data processing to training a machine learning model) for microbiome-based IBD diagnostic tools. We collected fifteen 16S rRNA microbiome datasets (7,707 samples) from North America to benchmark combinations of gut microbiome features, data normalization and transformation methods, batch effect correction methods, and machine learning models. Pipeline generalizability to new cohorts of patients was evaluated with two binary classification metrics following leave-one-dataset-out cross (LODO) validation, where all samples from one study were left out of the training set and tested upon. We demonstrate that taxonomic features processed with a compositional transformation method and batch effect correction with the naive zero-centering method attain the best classification performance. In addition, machine learning models that identify non-linear decision boundaries between labels are more generalizable than those that are linearly constrained. Lastly, we illustrate the importance of generating a curated training dataset to ensure similar performance across patient demographics. These findings will help improve the generalizability of machine learning models as we move towards non-invasive diagnostic and disease management tools for patients with IBD.

Integrative genomic analysis of PPP3R1 in Alzheimer's disease: a potential biomarker for predictive, preventive, and personalized medical approach

Alzheimer's disease (AD) is associated with abnormal calcium signaling, a pathway regulated by the calcium-dependent protein phosphatase. This study aimed to investigate the molecular function of protein phosphatase 3 regulatory subunit B (PPP3R1) underlying AD, which may provide novel insights for the predictive diagnostics, targeted prevention, and personalization of medical services in AD by targeting PPP3R1. A total of 1860 differentially expressed genes (DEGs) from 13,049 background genes were overlapped in AD/control and PPP3R1-low/high cohorts. Based on these DEGs, six co-expression modules were constructed by weight gene correlation network analysis (WGCNA). The turquoise module had the strongest correlation with AD and low PPP3R1, in which DEGs participated in axon guidance, glutamatergic synapse, long-term potentiation (LTP), mitogen-activated protein kinase (MAPK), Ras, and hypoxia-inducible factor 1 (HIF-1) signaling pathways. Furthermore, the cross-talking pathways of PPP3R1, such as axon guidance, glutamatergic synapse, LTP, and MAPK signaling pathways, were identified in the global regulatory network. The area under the curve (AUC) analysis showed that low PPP3R1 could accurately predict the onset of AD. Therefore, our findings highlight the involvement of PPP3R1 in the pathogenesis of AD via axon guidance, glutamatergic synapse, LTP, and MAPK signaling pathways, and identify downregulation of PPP3R1 as a potential biomarker for AD treatment in the context of 3P medicine-predictive diagnostics, targeted prevention, and personalization of medical services.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13167-021-00261-2.

Mathematical-based microbiome analytics for clinical translation

Traditionally, human microbiology has been strongly built on the laboratory focused culture of microbes isolated from human specimens in patients with acute or chronic infection. These approaches primarily view human disease through the lens of a single species and its relevant clinical setting however such approaches fail to account for the surrounding environment and wide microbial diversity that exists in vivo. Given the emergence of next generation sequencing technologies and advancing bioinformatic pipelines, researchers now have unprecedented capabilities to characterise the human microbiome in terms of its taxonomy, function, antibiotic resistance and even bacteriophages. Despite this, an analysis of microbial communities has largely been restricted to ordination, ecological measures, and discriminant taxa analysis. This is predominantly due to a lack of suitable computational tools to facilitate microbiome analytics. In this review, we first evaluate the key concerns related to the inherent structure of microbiome datasets which include its compositionality and batch effects. We describe the available and emerging analytical techniques including integrative analysis, machine learning, microbial association networks, topological data analysis (TDA) and mathematical modelling. We also present how these methods may translate to clinical settings including tools for implementation. Mathematical based analytics for microbiome analysis represents a promising avenue for clinical translation across a range of acute and chronic disease states.

Enhancing diversity analysis by repeatedly rarefying next generation sequencing data describing microbial communities

Amplicon sequencing has revolutionized our ability to study DNA collected from environmental samples by providing a rapid and sensitive technique for microbial community analysis that eliminates the challenges associated with lab cultivation and taxonomic identification through microscopy. In water resources management, it can be especially useful to evaluate ecosystem shifts in response to natural and anthropogenic landscape disturbances to signal potential water quality concerns, such as the detection of toxic cyanobacteria or pathogenic bacteria. Amplicon sequencing data consist of discrete counts of sequence reads, the sum of which is the library size. Groups of samples typically have different library sizes that are not representative of biological variation; library size normalization is required to meaningfully compare diversity between them. Rarefaction is a widely used normalization technique that involves the random subsampling of sequences from the initial sample library to a selected normalized library size. This process is often dismissed as statistically invalid because subsampling effectively discards a portion of the observed sequences, yet it remains prevalent in practice and the suitability of rarefying, relative to many other normalization approaches, for diversity analysis has been argued. Here, repeated rarefying is proposed as a tool to normalize library sizes for diversity analyses. This enables (i) proportionate representation of all observed sequences and (ii) characterization of the random variation introduced to diversity analyses by rarefying to a smaller library size shared by all samples. While many deterministic data transformations are not tailored to produce equal library sizes, repeatedly rarefying reflects the probabilistic process by which amplicon sequencing data are obtained as a representation of the amplified source microbial community. Specifically, it evaluates which data might have been obtained if a particular sample's library size had been smaller and allows graphical representation of the effects of this library size normalization process upon diversity analysis results.

Effects of dietary exposure to the engineered nanomaterials CeO2, SiO2, Ag, and TiO2 on the murine gut microbiome

Rodent studies on the effects of engineered nanomaterials (ENM) on the gut microbiome have revealed contradictory results. Our aim was to assess the effects of four well-investigated model ENM using a realistic exposure scenario. Two independent ad libitum feeding studies were performed. In study 1, female mice from the local breeding facility received feed pellets containing 1% CeO₂ or 1% SiO₂ for three weeks. In study 2, both female and male mice were purchased and exposed to 0.2% Ag-PVP or 1% TiO₂ for four weeks. A next generation 16S rDNA sequencing-based approach was applied to assess impacts on the gut microbiome. None of the ENM had an effect on the α- or β-diversity. A decreased relative abundance of the phylum Actinobacteria was observed in SiO₂ exposed mice. In female mice, the relative abundance of the genus Roseburia was increased with Ag exposure. Furthermore, in study 2, a sex-related difference in the β-diversity was observed. A difference in the β-diversity was also shown between the female control mice of the two studies. We did not find major effects on the gut microbiome. This contrast to other studies may be due to variations in the study design. Our investigation underlined the important role of the sex of test animals and their microbiome composition prior to ENM exposure initiation. Hence, standardization of microbiome studies is strongly required to increase comparability. The ENM-specific effects on Actinobacteria and Roseburia, two taxa pivotal for the human gut homeostasis, warrant further research on their relevance for health.

NetCoMi: network construction and comparison for microbiome data in R

MOTIVATION

Estimating microbial association networks from high-throughput sequencing data is a common exploratory data analysis approach aiming at understanding the complex interplay of microbial communities in their natural habitat. Statistical network estimation workflows comprise several analysis steps, including methods for zero handling, data normalization and computing microbial associations. Since microbial interactions are likely to change between conditions, e.g. between healthy individuals and patients, identifying network differences between groups is often an integral secondary analysis step. Thus far, however, no unifying computational tool is available that facilitates the whole analysis workflow of constructing, analysing and comparing microbial association networks from high-throughput sequencing data.

RESULTS

Here, we introduce NetCoMi (Network Construction and comparison for Microbiome data), an R package that integrates existing methods for each analysis step in a single reproducible computational workflow. The package offers functionality for constructing and analysing single microbial association networks as well as quantifying network differences. This enables insights into whether single taxa, groups of taxa or the overall network structure change between groups. NetCoMi also contains functionality for constructing differential networks, thus allowing to assess whether single pairs of taxa are differentially associated between two groups. Furthermore, NetCoMi facilitates the construction and analysis of dissimilarity networks of microbiome samples, enabling a high-level graphical summary of the heterogeneity of an entire microbiome sample collection. We illustrate NetCoMi's wide applicability using data sets from the GABRIELA study to compare microbial associations in settled dust from children's rooms between samples from two study centers (Ulm and Munich).

AVAILABILITY

R scripts used for producing the examples shown in this manuscript are provided as supplementary data. The NetCoMi package, together with a tutorial, is available at https://github.com/stefpeschel/NetCoMi.

CONTACT

Tel:+49 89 3187 43258; stefanie.peschel@mail.de.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Briefings in Bioinformatics online.