The target of inference in microbiome analyses is usually relative abundance (RA) because RA in a sample (e.g., stool) can be considered as an approximation of RA in an entire ecosystem (e.g., gut). However, inference on RA suffers from the fact that RA are calculated by dividing absolute abundances (AA) over the common denominator (CD), the summation of all AA (i.e., library size). Because of that, perturbation in one taxon will result in a change in the CD and thus cause false changes in RA of all other taxa, and those false changes could lead to false positive/negative findings. We propose a novel analysis approach (IFAA) to make robust inference on AA of an ecosystem that can circumvent the issues induced by the CD problem and compositional structure of RA. IFAA can also address the issues of overdispersion and handle zero-inflated data structures. IFAA identifies microbial taxa associated with the covariates in Phase one and estimates the association parameters by employing an independent reference taxon in Phase two. Two real data applications are presented and extensive simulations show that IFAA outperforms other established existing approaches by a big margin in the presence of unbalanced library size.

In order to test for the differential abundance of taxa that may drive the differences observed between inferred microbial communities derived from the different DNA isolation procedures, we performed DESeq2 analyses. Here we provide an example for such an analysis from human fecal specimen, examined using 16S rRNA gene profiling. This workflow relates to the article: Berith E. Knudsen, Lasse Bergmark, Patrick Munk, Oksana Lukjancenko, Anders Priemé, Frank M. Aarestrup, Sünje J. Pamp (2016) Impact of Sample Type and DNA Isolation Procedure on Genomic Inference of Microbiome Composition. mSystems Oct 2016, 1 (5) e00095-16; DOI: 10.1128/mSystems.00095-16

http://msystems.asm.org/content/1/5/e00095-16

Supplementary Table S2.5.4: ANCOM-BC analysis showing the differentially abundant predicted metabolic pathways (MetaCyc) in the amendment layer (1-2 cm beside the amendment) of planted soils without deep-banded amendment (DBA) compared to unplanted soils without DBA

This repository contains data files exported from BugSigDB, a manually curated database of published microbial signatures.

Release v1.3.0

• New signatures
• PubMed ID is the Study ID if available
• Taxonomic rank domain supported

The target of inference in microbiome analyses is usually relative abundance (RA) because RA in a sample (e.g., stool) can be considered as an approximation of RA in an entire ecosystem (e.g., gut). However, inference on RA suffers from the fact that RA are calculated by dividing absolute abundances (AAs) over the common denominator (CD), the summation of all AA (i.e., library size). Because of that, perturbation in one taxon will result in a change in the CD and thus cause false changes in RA of all other taxa, and those false changes could lead to false positive/negative findings. We propose a novel analysis approach (IFAA) to make robust inference on AA of an ecosystem that can circumvent the issues induced by the CD problem and compositional structure of RA. IFAA can also address the issues of overdispersion and handle zero-inflated data structures. IFAA identifies microbial taxa associated with the covariates in Phase 1 and estimates the association parameters by employing an independent reference taxon in Phase 2. Two real data applications are presented and extensive simulations show that IFAA outperforms other established existing approaches by a big margin in the presence of unbalanced library size. Supplementary materials for this article are available online.

Additional file 2: Tables S1-S7. Table S1- Correlation analysis between gut enterotypes and clinical factors. Table S2- Differential abundance analysis in endometriosis and control groups. Species with a prevalence > 10% and relative abundance ≥ 0.1% were compared in endometriosis and control groups using an Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC). Table S3- Differential abundance analysis in endometriosis and control groups. KEGG orthologs (KO) with a prevalence > 10% and relative abundance ≥ 0.1% were compared in endometriosis and control groups using an Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC). Table S4- Differential abundance analysis in endometriosis and control groups. EggNOG orthologs with a prevalence > 10% and relative abundance ≥ 0.1% were compared in endometriosis and control groups using an Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC). Table S5- Sensitivity differential abundance analysis in endometriosis and control groups. Species with a prevalence > 10% and relative abundance ≥ 0.1% were compared in endometriosis and control groups after excluding women with age > 50, using an Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC). Table S6- Sensitivity differential abundance analysis in endometriosis and control groups. KEGG orthologs (KO) with a prevalence > 10% and relative abundance ≥ 0.1% were compared in endometriosis and control groups after excluding women with age > 50, using an Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC). Table S7- Estrogen path enzymes with ALDEx2 analysis between endometriosis and control groups.

This repository contains data files exported from BugSigDB, a manually curated database of published microbial signatures.

Differential abundance analysis is a crucial task in many microbiome studies, where the central goal is to identify microbiome taxa associated with certain biological or clinical conditions. There are two different modes of microbiome differential abundance analysis: the individual-based univariate differential abundance analysis and the group-based multivariate differential abundance analysis. The univariate analysis identifies differentially abundant microbiome taxa subject to multiple correction under certain statistical error measurements such as false discovery rate, which is typically complicated by the high-dimensionality of taxa and complex correlation structure among taxa. The multivariate analysis evaluates the overall shift in the abundance of microbiome composition between two conditions, which provides useful preliminary differential information for the necessity of follow-up validation studies. In this paper, we present a novel Adaptive multivariate two-sample test for Microbiome Differential Analysis (AMDA) to examine whether the composition of a taxa-set are different between two conditions. Our simulation studies and real data applications demonstrated that the AMDA test was often more powerful than several competing methods while preserving the correct type I error rate. A free implementation of our AMDA method in R software is available at https://github.com/xyz5074/AMDA.

Additional file 6: Table S5. Dataset 1 false discovery rate (FDR) q-values for differential abundance methods when performed on filtered data. Dataset 1 FDR q-values for all differential abundance methods were aggregated into one table along with the mean relative abundance of each genus for PD patients (Case MRA) and control subjects (Control MRA) and the mean relative abundance ratio of Case MRA to Control MRA (MRAR). 201 PD patients and 132 controls were included in all analyses. Genera detected in at least 10% of samples (133 genera) were included in the analyses. If an analysis resulted in an "NA" for a result, or a result was not outputted by the method, a 1 was placed for the FDR q-value. Calculations for the number of detected DA signatures for each method are located at the bottom of the table. Calculations for the number of methods that detected or replicated a genus as differentially abundant are located to the right of the table.

Additional file 7: Table S6. Dataset 2 false discovery rate (FDR) q-values for differential abundance methods when performed on filtered data. Dataset 2 FDR q-values for all differential abundance methods were aggregated into one table along with the mean relative abundance of each genus for PD patients (Case MRA) and control subjects (Control MRA) and the mean relative abundance ratio of Case MRA to Control MRA (MRAR). 323 PD patients and 184 controls were included in all analyses. Genera detected in at least 10% of samples (195 genera) were included in the analyses. If an analysis resulted in an "NA" for a result, or a result was not outputted by the method, a 1 was placed for the FDR q-value. Calculations for the number of detected DA signatures for each method are located at the bottom of the table. Calculations for the number of methods that detected or replicated a genus as differentially abundant are located to the right of the table.

Additional file 4: Table S3. Dataset 1 false discovery rate (FDR) q-values for differential abundance methods when performed on unfiltered data. Dataset 1 FDR q-values for all differential abundance methods were aggregated into one table along with the mean relative abundance of each genus for PD patients (Case MRA) and control subjects (Control MRA) and the mean relative abundance ratio of Case MRA to Control MRA (MRAR). 201 PD patients and 132 controls were included in all analyses. All genera detected in dataset 1 (445 genera) were included in the analyses. If an analysis resulted in an "NA" for a result, or a result was not outputted by the method, a 1 was placed for the FDR q-value. Calculations for the number of detected DA signatures for each method are located at the bottom of the table. Calculations for the number of methods that detected or replicated a genus as differentially abundant are located to the right of the table.

Recent expansion of microbiome research has uncovered connections between resident microbial communities and blood parasite risk, establishing the potential for microbial disease treatments such as probiotics in the future. However, this field has largely focused on humans and model organisms, leaving much unknown about how microbial communities might directly or indirectly impact parasite infection in wild populations and non-mammals. To contribute to this knowledge base in wild birds, we collected fecal and blood samples from wild Eurasian tree sparrows (Passer montanus) in the United States to test for associations between blood parasite infection and the gut microbiome. We used a widespread molecular approach to test 81 samples from peripheral blood for Plasmodium and Haemoproteus, and we characterized the gut microbiome using fecal samples as a proxy. Neither alpha nor beta diversity significantly varied with detected Plasmodium infection. However, differential abundance analysis h...

Water chlorination is essential for controlling harmful microbes in drinking water; however, the antimicrobial effects of chlorine-based disinfectants may negatively impact the developing infant microbiota. This trial investigated the effects of chlorinated water on the infant gut microbiome. The waTer qUality and Microbiome Study (TUMS) was a double-blinded, randomised controlled trial. Six-month old infants (n=197) received either de-chlorinated drinking water via benchtop filtration (treatment, n=99), or regular chlorinated water (control, n=98) for twelve months. Tap water and stool samples were collected at baseline and at end of intervention. Metagenomic sequencing was used for faecal microbiome analysis. Primary outcomes were differences in gut microbiota alpha and beta diversity. Secondary outcomes included changes in the differential abundance of amplicon sequence variants (ASVs) and functional profiles. This study was registered with Australian New Zealand Clinical Trial Registry: ACTRN12619000458134.

Comparison of MetaDAVis and other popular microbiome analysis tools.

This dataset originates from a mouse model-based study in the field of gut microbial toxicology. It aims to investigate the effects of dietary emulsifiers—sodium carboxymethyl cellulose (CMC)—in combination with the organophosphate pesticide chlorpyrifos (CPF) on gut microbiota composition and function. The study further validates the mechanistic role of gut microbes via fecal microbiota transplantation (FMT) and sodium butyrate supplementation. All experiments were conducted in SPF-grade animal facilities at China Agricultural University. Detailed experimental methods are described in the associated manuscript.This dataset includes raw gut microbiota data from the following three components:Exposure experiment (The effects of food emulsifiers on chlorpyrifos-induced intestinal inflammation): Mice were administered different combinations of CPF and emulsifiers (CMC or soy lecithin, SL) across six treatment groups.Fecal microbiota transplantation (FMT): Fecal samples from exposed mice were transferred into microbiota-depleted mice (via antibiotic pretreatment) to assess functional consequences of altered microbial communities.Sodium butyrate supplementation: Butyrate was administered under CPF exposure conditions to evaluate its effect on restoring microbial and immune homeostasis.Microbial DNA was extracted from ileal and colonic contents. High-throughput sequencing of the 16S rRNA V3–V4 region was performed using the Illumina NovaSeq platform. Raw sequencing data were quality-filtered and trimmed using Fastp, followed by feature table construction via QIIME2 (v2022.2) using the DADA2 pipeline. Taxonomic annotation was carried out using the SILVA 138 database. All data were processed and exported using the Majorbio Cloud Platform (https://cloud.majorbio.com/).All sequencing samples passed quality control. A few low-depth or contaminated samples were removed and are not included in the dataset. No systematic missing data are present. The data are unrarefied and suitable for α/β diversity analysis, differential abundance analysis, and microbial community structure assessment.This dataset provides foundational information for exploring how co-exposure to food additives and environmental toxicants disrupts gut microbiota composition and metabolism. It also supports downstream ecological, toxicological, and metabolic network analyses.

The lower airway microbiota in patients with chronic obstructive pulmonary disease (COPD) are likely altered compared with the microbiota in healthy individuals. Information on how the microbiota is affected by smoking, the use of inhaled corticosteroids (ICS), and COPD severity is still scarce. In the MicroCOPD Study, participant characteristics were obtained through standardised questionnaires and clinical measurements at a single centre from 2012 to 2015. Protected bronchoalveolar lavage samples from 97 patients with COPD and 97 controls were paired-end sequenced with the Illumina MiSeq System. Data were analysed in QIIME 2 and R. Alpha-diversity was lower in patients with COPD than controls (Pielou evenness: COPD=0.76, control=0.80, p=0.004; Shannon entropy: COPD=3.98, control=4.34, p=0.01). Beta-diversity differed with smoking only in the COPD cohort (weighted UniFrac: permutational analysis of variance R2=0.04, p=0.03). Nine genera were differentially abundant between COPD and controls. Genera enriched in COPD belonged to the Firmicutes phylum. Pack years were linked to differential abundance of taxa in controls only (ANCOM-BC (Analysis of Compositions of Microbiomes with Bias Correction) log-fold difference/q-values: Haemophilus -0.05/0.048; Lachnoanaerobaculum -0.04/0.03). Oribacterium was absent in smoking patients with COPD compared with non-smoking patients (ANCOM-BC log-fold difference/q-values: -1.46/0.03). We found no associations between the microbiota and COPD severity or ICS. The lower airway microbiota is equal in richness in patients with COPD to controls, but less even. Genera from the Firmicutes phylum thrive particularly in COPD airways. Smoking has different effects on diversity and taxonomic abundance in patients with COPD compared with controls. COPD severity and ICS use were not linked to the lower airway microbiota.

Acute hepatopancreatic necrosis disease (AHPND) is an emerging bacterial disease of cultured shrimp caused mainly by Vibrio parahaemolyticus, which harbors the lethal PirAB toxin genes. Although Penaeus vannamei (P. vannamei) postlarvae are susceptible to AHPND, the changes in the bacterial communities through the larval stages affected by the disease are unknown. We characterized, through high-throughput sequencing, the microbiome of P. vannamei larvae infected with AHPND-causing bacteria through the larval stages and compared the microbiome of larvae collected from high- and low-survival tanks. A total of 64 tanks from a commercial hatchery were sampled at mysis 3, postlarvae 4, postlarvae 7, and postlarvae 10 stages. PirAB toxin genes were detected by PCR and confirmed by histopathology analysis in 58 tanks. Seven from the 58 AHPND-positive tanks exhibited a survival rate higher than 60% at harvest, despite the AHPND affectation, being selected for further analysis, whereas 51 tanks exhibited survival rates lower than 60%. A random sample of 7 out of these 51 AHPND-positive tanks was also selected. Samples collected from the selected tanks were processed for the microbiome analysis. The V3–V4 hypervariable regions of the 16S ribosomal RNA (rRNA) gene of the samples collected from both the groups were sequenced. The Shannon diversity index was significantly lower at the low-survival tanks. The microbiomes were significantly different between high- and low-survival tanks at M3, PL4, PL7, but not at PL10. Differential abundance analysis determined that biomarkers associated with high and low survival in shrimp hatchery tanks affected with AHPND. The genera Bacillus, Vibrio, Yangia, Roseobacter, Tenacibaculum, Bdellovibrio, Mameliella, and Cognatishimia, among others, were enriched in the high-survival tanks. On the other hand, Gilvibacter, Marinibacterium, Spongiimonas, Catenococcus, and Sneathiella, among others, were enriched in the low-survival tanks. The results can be used to develop applications to prevent losses in shrimp hatchery tanks affected by AHPND.

Background. Recent evidence suggests that multigenic and complex environmentally modulated diseases result from early life exposure to toxicants at least partly via gut microbial influences. Environmental toxicants, polybrominated diphenyl ethers (PBDEs), and polychlorinated biphenyls (PCBs) are breast milk-enriched persistent organic pollutants (POPs) and thus remain a continuing threat to human health despite being banned from production. Recent findings focused on the liver developmental reprogramming capabilities from neonatal BPA exposure; however, little is known on how PBDEs and PCBs regulate the liver transcriptome with respect to the gut microbiome.

Objectives. We investigated whether the gut microbiome can be persistently reprogrammed with the liver following neonatal exposure to POPs, and whether microbial biomarkers associated with disease-prone changes in the hepatic epigenetic and transcriptomic landscape in adulthood.

Methods. C57BL/6 male and female mouse pups were orally administered vehicle, bisphenol A (BPA), BDE-99 (a breast milk-enriched PBDE congener), or the Fox River PCB mixture (an environmentally relevant PCB mixture), between postnatal day (PND) 2 to 4, once daily for three consecutive days. Tissues were collected at PND5 and PND60 for 16S rDNA sequencing and targeted metabolomics.

Results. Neonatal exposure to BDE-99, followed by BPA and PCB, produced the greatest persistent changes in the adult hepatic transcriptome, including an inflammation and cancer-prone transcriptomic signature. BDE-99 exposure resulted in a persistent increase in Akkermansia muciniphila throughout the intestinal sections and feces. We observed persistent increases in acetate and succinate, metabolites A. muciniphila is able to produce. Correspondingly, liver H3K4me1 and H3K27 acetylation were enriched around the loci encoding liver cancer-related genes following neonatal BDE-99 exposure.

Conclusion. Similar to BPA, early life exposure to BDE-99 also produced a cancer-prone hepatic transcriptomic signature corresponding to an increase in permissive epigenetic signatures around cancer-related genes in adulthood. This positively associates with BDE-99 mediated increase in A. muciniphila and its metabolites which are established epigenetic modifiers.

Methods Chemicals. 2,2’,4,4’,5-pentabromodiphenyl-ether (BDE-99) (CAS No. 60348-60-9) and bisphenol A (BPA) were purchased from AccuStandard, Inc. (New Haven, CT). Fox River PCB mixture was prepared from technical PCB mixtures as we described previously (Lim et al. 2020). Corn oil was purchased from Sigma-Aldrich (St. Louis, MO). All other chemicals, unless otherwise noted, were purchased from Sigma-Aldrich St. Louis, MO).

Animals

All mice were housed according to the Association for Assessment and Accreditation of Laboratory Animal Care International guidelines (https://aaalac.org/resources/theguide.cfm), and studies were approved by the Institutional Animal Care and Use Committee at the University of Washington. Eight-week-old specific pathogen-free C57BL/6J mice were purchased from the Jackson Laboratory (Bar Harbor, ME) and were acclimated to the animal facility at the University of Washington for at least two breeding generations prior to the experiment. Mice were housed in standard air-filtered cages using autoclaved bedding (autoclaved Enrich-N’Pure) (Andersons, Maumee, OH), and were exposed ad libitum to non-acidified autoclaved water, as well as LabDiet #5021 for breeding pairs, or LabDiet #5010 for post-weaning pups (LabDiet, St. Louis, MO). All chemical solutions were sterilized using the Sterflip Vacuum-Driven Filtration System with a 0.22 μm Millipore Express Plus Membrane (EMD Millipore, Temecula, CA). As shown in the study design diagram (Fig. 1A), litters were randomly assigned to chemical or vehicle exposures between postnatal day (PND) 2 to 4, pups were supralingually exposed to corn oil (vehicle, 5 ml/kg), BPA (250 μg/kg), BDE-99 (57 mg/kg), or the PCB Fox River Mixture (30 mg/kg), once daily for three consecutive days (n= 5/sex/exposure). Mice were euthanized using CO2 and decapitation (PND5), or CO2 followed by exsanguination via cardiac puncture (PND60). Various bio-compartments were collected at PND5 (24 hours after the exposure) and 60 (56 days after the last exposure), including small and large intestine sections (duodenum, jejunum, ileum, and colon), feces, liver, and serum. Samples were stored in a -80˚C freezer until further analysis as described in Fig. 1A and detailed below: 16S rDNA sequencing and data analysis Microbial DNA was isolated from duodenum, jejunum, ileum, colon, and feces at PND5 and 60, as well as liver and serum at PND60, using an OMEGA E.Z.N.A. Stool DNA Kit (OMEGA Biotech Inc., Norcross, Georgia). To note, the microbial DNA in liver and serum serves as an indirect biomarker for gut permeability. The concentration of DNA was determined using Qubit (Thermo Fisher Scientific, Waltham, Massachusetts). The integrity and quantity of all DNA samples were confirmed using an Agilent 2100 Bioanalyzer (Agilent Technologies Inc., Santa Clara, California). Bacterial 16S rDNA hypervariable region 4 (V4) amplicon sequencing was performed using a HiSeq 2500 second-generation sequencer (250 bp paired-end) (n = 5 per group) using a similar method as we described before (Cheng et al. 2018; Gomez et al. 2021; Lim et al. 2020; Scoville et al. 2019). Microbial 16S rDNA paired-end sequence reads, in quintuplicate, were merged, de-multiplexed, quality-checked, and chimera-filtered using QIIME 2 (https://qiime2.org/)(Bolyen et al. 2019). The Silva 99 version 132 reference was used (Quast et al. 2013). The operational taxonomic unit (OTU) table annotated with classified bacteria information was read into R for further analysis (R Core Team, 2019). Differential abundance testing was performed using the ANCOM plugin in QIIME 2 (Mandal et al. 2015). Differentially abundant bacteria (relative abundance of OTU) and differentially regulated hepatic genes were correlated (Pearson’s correlation), and significant correlation was defined as |r| > 0.8 and false discovery rate (FDR)-adjusted p-value < 0.1. Correlation results were plotted using ComplexHeatmap (Gu et al. 2016). Functional predictions of the microbial composition were conducted using PICRUSt2 (Douglas et al. 2020).

RNA Isolation

Total RNA was isolated from frozen livers using the RNA-Bee reagent (Tel-Test lnc., Friendswood, TX) according to the manufacturer’s protocol. RNA concentrations were quantified using a NanoDrop 1000 Spectrophotometer (Thermo Scientific, Waltham, MA) at 260 nm. The integrity of total RNA samples was evaluated by formaldehyde-agarose gel electrophoresis with visualization of 18S and 28S rRNA bands under UV light and confirmed by an Agilent 2100 Bioanalyzer (Agilent Technologies Inc., Santa Clara, CA). Samples with RNA integrity (RIN) values above 8.0 were used for RNA-Seq.

Whole transcriptome Sequencing of liver samples. In triplicates, the cDNA library was constructed using a ribosomal depletion method, and reads were sequenced (75 bp paired end) per the Illumina manufacturer’s protocol. FASTQ files were de-multiplexed and concatenated for each sample. Quality control the FASTQ files was performed using FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Reads were then mapped to the mouse reference genome (National Center for Biotechnology Information [NCBI GRCm38/mm10]) using HISAT2 version 2.1 (Kim et al. 2015). The sequencing alignment/map (SAM) files were converted to binary alignment/map (BAM) format using SAMtools version 1.8 (Li et al., 2009) and were analyzed by Cufflinks version 2.2.1 to estimate the transcript abundance (Trapnell et al. 2010) using Gencode mouse version 19 (vM19) gene transfer format (GTF). Long non-coding RNA (lncRNA) GTF from Gencode (vM19) was used to estimate the lncRNA transcript abundance. The abundance was expressed as fragments per kilobase of transcript per million mapped reads (FPKM) which was then converted to transcripts per million (TPM). Differential expression analysis was performed using Cuffdiff (Trapnell et al. 2010). The differentially expressed genes were defined as false discovery rate Benjamini-Hotchberg adjusted p value (FDR-BH) <0.05 in the chemical-exposed groups compared with the vehicle-exposed control group. RNA-Seq data analysis. All analyses were done in R unless stated otherwise. For RNA-Seq, samples were categorized by age, sex, and exposure. Genes were considered expressed if the average TPM was >1 for at least one group. Protein coding genes (PCGs) were separately categorized based on the Ensembl BioMart gene category. Differentially expressed PCGs were also categorized and matched with genes in categories of interest (xenobiotic biotransformation, epigenetic modifiers, nuclear receptors, and oxidative stress). These specific gene categories were based on literature searches and extracting gene sets from the Gene Ontology consortium and the Kyoto Encyclopedia of Genes and Genomes (KEGG). Venn diagrams were made for differentially regulated PCGs and lncRNAs for each sex and age group (PND5 and 60) using the R package VennDiagram (Chen and Boutros 2011). Hierarchical clustering was generated clustering using R package ComplexHeatmap (Gu et al. 2016). Up- and down-regulations were defined as the absolute value of the fold change greater than 1.5. Lists of up- and down-regulated genes were used as input for gene ontology enrichment using the R package topGO (Alexa and Rahnenfuhrer, 2020) for all groups. The list of genes in the unfiltered expression table was used as the background for gene ontology. Upstream regulators for all differentially regulated genes were determined using IPA. Gene expression values were plotted using bar plots for specific genes of interest using the arithmetic mean TPM in SigmaPlot.

Chromatin

Fecal microbial biomarkers represent a less invasive alternative for acquiring information on wildlife populations than many traditional sampling methodologies. Our goal was to evaluate linkages between fecal microbiome communities in Rocky Mountain elk (Cervus canadensis) and four host factors including sex, age, population, and physical condition (body-fat). We paired a feature-selection algorithm with an LDA-classifier trained on elk differential bacterial abundance (16S-rRNA amplicon survey) to predict host health factors from 104 elk microbiomes across four elk populations. We validated the accuracy of the various classifier predictions with leave-one-out cross-validation using known measurements. We demonstrate that the elk fecal microbiome can predict the four host factors tested. Our results show that elk microbiomes respond to both the strong extrinsic factor of biogeography and simultaneously occurring, but more subtle, intrinsic forces of individual body-fat, sex, and age class. Thus, we have developed and described herein a generalizable approach to disentangle microbiome responses attributed to multiple host factors of varying strength from the same bacterial sequence data set. Wildlife conservation and management presents many challenges, but we demonstrate that non-invasive microbiome surveys from scat samples can provide alternative options for wildlife population monitoring. We believe that, with further validation, this method could be broadly applicable in other species and potentially predict other measurements. Our study can help guide the future development of microbiome-based monitoring of wildlife populations and supports hypothetical expectations found in host-microbiome theory.

Differential abundance analysis between DT and CFC sample groups using EdgeR.