Longitudinal, prospective studies often rely on multi-omics approaches, wherein various specimens are analyzed for genomic, metabolomic, and/or transcriptomic profiles. In practice, longitudinal studies in humans and other animals routinely suffer from subject dropout, irregular sampling, and biological variation that may not be normally distributed. As a result, testing hypotheses about observations over time can be statistically challenging without performing transformations and dramatic simplifications to the dataset, causing a loss of longitudinal power in the process. Here, we introduce splinectomeR, an R package that uses smoothing splines to summarize data for straightforward hypothesis testing in longitudinal studies. The package is open-source, and can be used interactively within R or run from the command line as a standalone tool. We present a novel in-depth analysis of a published large-scale microbiome study as an example of its utility in straightforward testing of key hypotheses. We expect that splinectomeR will be a useful tool for hypothesis testing in longitudinal microbiome studies.

Output files of the application of our R software (available at https://github.com/wilkinsonlab/robust-clustering-metagenomics) to different microbiome datasets already published.

Prefixes:

• David2014_: original microbiome dataset published in [David et al.,2014] (http://genomebiology.com/2014/15/7/R89)
• Ballou2016_: original microbiome dataset published in [Ballou et al.,2016] (http://journal.frontiersin.org/article/10.3389/fvets.2016.00002/full)
• Gajer2012_: original microbiome dataset published in [Gajer et al.,2012] (http://stm.sciencemag.org/content/4/132/132ra52.long)
• LaRosa2014_: original microbiome dataset published in [LaRosa et al.,2014] (http://www.pnas.org/cgi/doi/10.1073/pnas.1409497111)

Suffixes:

• _All: all taxa

• _Dominant: only 1% most abundant taxa

• _NonDominant: remaining taxa after removing above dominant taxa

• _GenusAll: taxa aggregated at genus level

• _GenusDominant: taxa aggregated at genes level and then to select only 1% most abundant taxa

• _GenusNonDominant: taxa aggregated at genus level and then to remove 1% most abundant taxa

Each folder contains 3 output files related to the same input dataset:
- data.normAndDist_definitiveClustering_XXX.RData: R data file with a) a phyloseq object (including OTU table, meta-data and cluster assigned to each sample); and b) a distance matrix object.
- definitiveClusteringResults_XXX.txt: text file with assessment measures of the selected clustering.
- sampleId-cluster_pairs_XXX.txt: text file. Two columns, comma separated file: sampleID,clusterID

Abstract of the associated paper:

The analysis of microbiome dynamics would allow us to elucidate patterns within microbial community evolution; however, microbiome state-transition dynamics have been scarcely studied. This is in part because a necessary first-step in such analyses has not been well-defined: how to deterministically describe a microbiome's "state". Clustering in states have been widely studied, although no standard has been concluded yet. We propose a generic, domain-independent and automatic procedure to determine a reliable set of microbiome sub-states within a specific dataset, and with respect to the conditions of the study. The robustness of sub-state identification is established by the combination of diverse techniques for stable cluster verification. We reuse four distinct longitudinal microbiome datasets to demonstrate the broad applicability of our method, analysing results with different taxa subset allowing to adjust it depending on the application goal, and showing that the methodology provides a set of robust sub-states to examine in downstream studies about dynamics in microbiome.

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

Simultaneous profiling of biospecimens using different technological platforms enables the study of many data types, encompassing microbial communities, omics, and meta-omics as well as clinical or chemistry variables. Reduction in costs now enables longitudinal or time course studies on the same biological material or system. The overall aim of such studies is to investigate relationships between these longitudinal measures in a holistic manner to further decipher the link between molecular mechanisms and microbial community structures, or host-microbiota interactions. However, analytical frameworks enabling an integrated analysis between microbial communities and other types of biological, clinical, or phenotypic data are still in their infancy. The challenges include few time points that may be unevenly spaced and unmatched between different data types, a small number of unique individual biospecimens, and high individual variability. Those challenges are further exacerbated by the inherent characteristics of microbial communities-derived data (e.g., sparse, compositional). We propose a generic data-driven framework to integrate different types of longitudinal data measured on the same biological specimens with microbial community data and select key temporal features with strong associations within the same sample group. The framework ranges from filtering and modeling to integration using smoothing splines and multivariate dimension reduction methods to address some of the analytical challenges of microbiome-derived data. We illustrate our framework on different types of multi-omics case studies in bioreactor experiments as well as human studies.

BackgroundStudies of colonoscopic fecal microbiota transplant (FMT) in patients with recurrent CDI, indicate that this is a very effective treatment for preventing further relapses. In order to provide this service at Stony Brook University Hospital, we initiated an open-label prospective study of single colonoscopic FMT among patients with ≥ 2 recurrences of CDI, with the intention of monitoring microbial composition in the recipient before and after FMT, as compared with their respective donor. We also initiated a concurrent open label prospective trial of single colonoscopic FMT of patients with ulcerative colitis (UC) not responsive to therapy, after obtaining an IND permit (IND 15642). To characterize how FMT alters the fecal microbiota in patients with recurrent Clostridia difficile infections (CDI) and/or UC, we report the results of a pilot microbiome analysis of 11 recipients with a history of 2 or more recurrences of C. difficile infections without inflammatory bowel disease (CDI-only), 3 UC recipients with recurrent C. difficile infections (CDI + UC), and 5 UC recipients without a history of C. difficile infections (UC-only).MethodV3V4 Illumina 16S ribosomal RNA (rRNA) gene sequencing was performed on the pre-FMT, 1-week post-FMT, and 3-months post-FMT recipient fecal samples along with those collected from the healthy donors. Fitted linear mixed models were used to examine the effects of Group (CDI-only, CDI + UC, UC-only), timing of FMT (Donor, pre-FMT, 1-week post-FMT, 3-months post-FMT) and first order Group*FMT interactions on the diversity and composition of fecal microbiota. Pairwise comparisons were then carried out on the recipient vs. donor and between the pre-FMT, 1-week post-FMT, and 3-months post-FMT recipient samples within each group.ResultsSignificant effects of FMT on overall microbiota composition (e.g., beta diversity) were observed for the CDI-only and CDI + UC groups. Marked decreases in the relative abundances of the strictly anaerobic Bacteroidetes phylum, and two Firmicutes sub-phyla associated with butyrate production (Ruminococcaceae and Lachnospiraceae) were observed between the CDI-only and CDI + UC recipient groups. There were corresponding increases in the microaerophilic Proteobacteria phylum and the Firmicutes/Bacilli group in the CDI-only and CDI + UC recipient groups. At a more granular level, significant effects of FMT were observed for 81 genus-level operational taxonomic units (OTUs) in at least one of the three recipient groups (p<0.00016 with Bonferroni correction). Pairwise comparisons of the estimated pre-FMT recipient/donor relative abundance ratios identified 6 Gammaproteobacteria OTUs, including the Escherichia-Shigella genus, and 2 Fusobacteria OTUs with significantly increased relative abundance in the pre-FMT samples of all three recipient groups (FDR < 0.05), however the magnitude of the fold change was much larger in the CDI-only and CDI + UC recipients than in the UC-only recipients. Depletion of butyrate producing OTUs, such as Faecalibacterium, in the CDI-only and CDI + UC recipients, were restored after FMT.ConclusionThe results from this pilot study suggest that the microbial imbalances in the CDI + UC recipients more closely resemble those of the CDI-only recipients than the UC-only recipients.

IntroductionRepeated measures microbiome studies, including longitudinal and clustered designs, offer valuable insights into the dynamics of microbial communities and their associations with various health outcomes. However, visualizing such multivariate data poses significant challenges, particularly in distinguishing meaningful biological patterns from noise introduced by covariates and the complexities of repeated measures.MethodsIn this study, we propose a framework to enhance the visualization of repeated measures microbiome data using Principal Coordinates Analysis (PCoA) adjusted for covariates through linear mixed models (LMM). Our method adjusts for confounding variables and accounts for the repeated measures structure of the data, enabling clearer identification of microbial community variations across time points or clusters.ResultsWe demonstrate the utility of our approach through simulated scenarios and real datasets, showing that it effectively mitigates the influence of nuisance covariates and highlights key axes of microbiome variation.DiscussionThis refined visualization technique provides a robust tool for researchers to explore and understand microbial community dynamics in repeated measures microbiome studies.

Stunting or reduced linear growth is very prevalent in low-income countries. Recent studies have demonstrated a causal relationship between alterations in the gut microbiome and moderate or severe acute malnutrition in children in these countries. However, there have been no primary longitudinal studies comparing the intestinal microbiota of persistently stunted children to that of non-stunted children in the same community. In this pilot study, we characterized gut microbial community composition and diversity of the fecal microbiota of 10 children with low birth weight and persistent stunting (cases) and 10 children with normal birth weight and no stunting (controls) from a birth cohort every 3 months up to 2 years of age in a slum community in south India. There was an increase in diversity indices (P <0.0001) with increasing age in all children. However, there were no differences in diversity indices or in the rates of their increase with increasing age between cases and controls. The percent relative abundance of the Bacteroidetes phylum was higher in stunted compared to control children at 12 months of age (P = 0.043). There was an increase in the relative abundance of this phylum with increasing age in all children (P = 0.0380) with no difference in the rate of increase between cases and controls. There was a decrease in the relative abundance of Proteobacteria (P = 0.0004) and Actinobacteria (P = 0.0489) with increasing age in cases. The microbiota of control children was enriched in probiotic species Bifidobacterium longum and Lactobacillus mucosae, whereas that of stunted children was enriched in inflammogenic taxa including those in the Desulfovibrio genus and Campylobacterales order. Larger, longitudinal studies on the compositional and functional maturation of the microbiome in children are needed.

No special programs of software are required to open the data files. The R scripts that were used to analyze the data can be found on GitHub (https://github.com/sjbd1/5xfAD_mBio2022). PRJNA902000 is the bioaccession number for the raw sequence reads on the SRA: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA902000/

Chickens are a key food source for humans yet their microbiome contains bacteria that can be pathogenic to humans, and indeed potentially to chickens themselves. Campylobacter is present within the chicken gut and is the leading cause of bacterial foodborne gastroenteritis within humans worldwide. Infection can lead to secondary sequelae such as Guillain-Barré syndrome and stunted growth in children from low-resource areas. Despite the global health impact and economic burden of Campylobacter, how and when Campylobacter appears within chickens remains unclear. The lack of day to day microbiome data with replicates, relevant metadata, and a lack of natural infection studies have delayed our understanding of the chicken gut microbiome and Campylobacter. Here, we performed a comprehensive day to day microbiome analysis of the chicken cecum from day 3 to 35 (12 replicates each day; final n = 379). We combined metadata such as chicken weight and feed conversion rates to investigate what the driving forces are for the microbial changes within the chicken gut over time, and how this relates to Campylobacter appearance within a natural habitat setting. We found a rapidly increasing microbial diversity up to day 12 with variation observed both in terms of genera and abundance, before a stabilization of the microbial diversity after day 20. In particular, we identified a shift from competitive to environmental drivers of microbial community from days 12 to 20 creating a window of opportunity whereby Campylobacter can appear. Campylobacter was identified at day 16 which was 1 day after the most substantial changes in metabolic profiles observed. In addition, microbial variation over time is most likely influenced by the diet of the chickens whereby significant shifts in OTU abundances and beta dispersion of samples often corresponded with changes in feed. This study is unique in comparison to the most recent studies as neither sampling was sporadic nor Campylobacter was artificially introduced, thus the experiments were performed in a natural setting. We believe that our findings can be useful for future intervention strategies and help reduce the burden of Campylobacter within the food chain.

A number of factors contribute to the complex array of small molecules that occur in stool; including diet, gut flora, and gut function. Comprehensive profiling of the stool metabolome therefore can provide detailed phenotypic information on health status, metabolic interactions between the host and the microbiome, and interactions among gut microbes. Here, we applied metabolomics to characterize stool samples collected longitudinally from inflammatory bowel disease (IBD) patients and non-IBD controls who participated in the Integrative Human Microbiome Project (iHMP). A total of 546 stool samples were analyzed using a platform comprised of four complementary liquid chromatography tandem mass spectrometry (LC-MS) methods designed to measure polar metabolites and lipids. Each method used high resolution/accurate mass (HRAM) profiling to measure both metabolites of confirmed identity and yet to be identified metabolite peaks. 81,867 de-isotoped LC-MS peaks were measured, out of which 597 were annotated based on confirmation with authentic reference standards. Pooled stool extracts inserted and analyzed throughout the analysis queues to evaluate analytical reproducibility showed a median coefficient of variation of 5.1% among known metabolites and 24.2% across all 81,867 features. Owing to differences in water content and heterogeneity among stool samples, total median scaling was used to standardize the metabolomics data. In addition to being accessible at the Metabolomics Workbench repository, these metabolomics data will be incorporated into a multi’omic database (https://www.hmpdacc.org/ihmp/) that will enable the study of associations between the gut microbiome and IBD.

Microbiome Virulome Profiling Market Outlook

According to our latest research, the global microbiome virulome profiling market size reached USD 1.12 billion in 2024. Driven by rapid advancements in sequencing technologies and the increasing significance of microbiome research in healthcare, the market is poised to expand at a CAGR of 14.6% from 2025 to 2033. By the end of the forecast period, the market is projected to attain a value of USD 3.53 billion in 2033. The primary growth factor for this robust expansion is the rising demand for precision medicine and the growing recognition of the role of microbial virulence factors in disease pathogenesis.

One of the most significant growth drivers in the microbiome virulome profiling market is the increasing adoption of next-generation sequencing (NGS) and advanced bioinformatics tools. These technologies have revolutionized the ability to analyze complex microbial communities and their virulence factors at an unprecedented scale and resolution. As healthcare systems worldwide prioritize personalized medicine, there is a heightened demand for comprehensive profiling of the human microbiome to better understand disease mechanisms, predict disease risk, and tailor therapeutic interventions. Additionally, the integration of artificial intelligence and machine learning into bioinformatics platforms is enhancing the accuracy and speed of data analysis, further accelerating market growth.

Another key factor propelling the market is the expanding application of microbiome virulome profiling in clinical diagnostics and drug discovery. By identifying specific virulence genes and pathogenic potential within microbial communities, clinicians can make more informed decisions regarding patient management, infection control, and targeted therapies. Pharmaceutical and biotechnology companies are also leveraging virulome data to develop novel therapeutics, probiotics, and vaccines aimed at modulating the microbiome for improved health outcomes. The increasing prevalence of infectious diseases, antibiotic resistance, and chronic conditions linked to dysbiosis underscores the critical need for advanced profiling solutions, thereby driving sustained investment and innovation in this sector.

The market is further supported by growing research initiatives and public-private partnerships focused on microbiome studies. Governments and academic institutions are investing heavily in microbiome research consortia, biobanking, and longitudinal studies to unravel the complex interactions between host and microbial virulence factors. These efforts are fostering a collaborative ecosystem that encourages the development of standardized protocols, robust databases, and translational research, all of which are essential for the wider adoption of virulome profiling in both clinical and research settings. The convergence of regulatory support, funding, and technological advancements is creating a fertile environment for market expansion.

Regionally, North America dominates the microbiome virulome profiling market, owing to its well-established healthcare infrastructure, significant R&D investments, and the presence of leading industry players. Europe follows closely, benefiting from strong government support for genomics research and a growing emphasis on preventive healthcare. The Asia Pacific region is emerging as a high-growth market, driven by increasing healthcare expenditure, rising awareness of microbiome-related diseases, and expanding capabilities in genomics and molecular diagnostics. Latin America and the Middle East & Africa are also witnessing gradual adoption, supported by improvements in healthcare access and international collaborations.

Technology Analysis

The technology segment of the microbiome virulome profiling market is primarily categorized into sequencing, PCR, microarray, and bioinformatics. Sequencing technologies, especially next-generation sequencing (NGS), have become the cornerstone of microbiome analysis due to their ability to deliver high-throughput, comprehensive data on microbial communities and their virulence factors. The adoption of NGS has dramatically reduced the cost and time required for profiling, making it accessible to a broader range of researchers and clinicians. As a result, sequencing accounted for the largest share of the technology segment in 2024, and this dominance is expected to persist throughout the forecast period. The continuous evolution of sequencing platforms, such as the introduction of third

Research has illustrated the presence of a diverse range of microbiota in human milk. The composition of the milk microbiome varies across different stages of lactation, emphasizing the need to consider the lactation stage when studying its composition. Additionally, the transfer of both milk and skin microbiota during breastfeeding is crucial for understanding their collective impact on infant health and development. Further exploration of the complete breastfeeding microbiome is necessary to unravel the role these organisms play in infant development. We aim to longitudinally assess the bacterial breastfeeding microbiome across stages of lactation. This includes all the bacteria that infants are exposed to during breastfeeding, such as bacteria found within human milk and any bacteria found on the breast and nipple. Forty-six human milk samples were collected from 15 women at 1, 4, 7, and 10 months postpartum. Metagenomic analysis of the bacterial microbiome for these samples was perf..., Participants Samples were collected from 15 women, who had experienced healthy pregnancies and deliveries, across lactation as part of the Baby Connectome Project and the Baby Connectome Project—Enriched, a joint effort between the University of North Carolina at Chapel Hill and the University of Minnesota Twin Cities. No women included reported having taken any antibiotics within 3 months of providing samples. Nine of the women exclusively breastfed through six months, while the other six did supplement with infant formula they still received more breast milk compared to infant formula through six months. Milk collection and processing Milk was collected at the University of Minnesota when the dyad was on site for behavioral data collection, and was timed to coincide with the 2nd feed of the day whenever possible. Each participant was provided a quiet, private space equipped with a Medela Symphony hospital grade breast pump and a sterilized set of pump consumables. Mothers were asked t..., , # Bacterial composition of the human milk microbiome across lactation

https://doi.org/10.5061/dryad.rfj6q57j9

Longitudinal bacterial composition of human milk across lactation in 15 healthy lactating people as analyzed in a forthcoming publication.

Description of the data and file structure

Data are organized into two files, Milk_Metagenome_Taxa_Table.csv and Sample_Age_Participant.txt.Â

Column definitions for Milk_Metagenome_Taxa_Table.csv:

   Name: Name of bacteria specified by CosmosID

   Kingdom: Taxonomic kingdom

   Phylum: Taxonomic phylum

   Class: Taxonomic class

   Order: Taxonomic order

   Family: Taxonomic family

   Genus: Taxonomic genus

   Species: Taxonomic species

   The 50 following column headers represent individual samples

   Taxonomy ID: ID assigned by CosmosID

   Taxonomic Lineage: Combined taxonomic lineage

Column definitions for Sample_Age_Participant.csv:

   Age (months): Infa...

Companion animals provide an excellent model for studies of the gut microbiome because potential confounders such as diet and environment can be more readily controlled for than in humans. Additionally, domestic cats and dogs are typically neutered early in life, enabling an investigation into the potential effect of sex hormones on the microbiome. In a longitudinal study to investigate the potential effects of neutering, neutering age and gender on the gut microbiome during growth, the faeces of kittens (16 male, 14 female) were sampled at 18, 30 and 42 weeks of age. DNA was shotgun sequenced on the Illumina platform and sequence reads were annotated for taxonomy and function by comparison to a database of protein coding genes. In a statistical analysis of diversity, taxonomy and functional potential of the microbiomes, age was identified as the only factor with significant associations. No significant effects were detected for gender, neutering, or age when neutered (19 or 31 weeks). At 18 weeks of age the microbiome was dominated by the genera Lactobacillus and Bifidobacterium (35% and 20% average abundance). Structural and functional diversity was significantly increased by week 30 but there was no further significant increase. At 42 weeks of age the most abundant genera were Bacteroides (16%), Prevotella (14%) and Megasphaera (8%). Significant differences in functional potential included an enrichment for genes in energy metabolism (carbon metabolism and oxidative phosphorylation) and depletion in cell motility (flagella and chemotaxis). We conclude that the feline faecal microbiome is predominantly determined by age when diet and environment are controlled for. We suggest this finding may also be informative for studies of the human microbiome, where control over such factors is usually limited.

Several studies have shown that the gut microbiota is associated with the brain structure. However, the causal relationship between the gut microbiota and structural changes in the brain remains undetermined. Gut microbiota data from genome-wide association (GWAS) meta-analyses summarized by the MiBioGen consortium were used (n=13,266) and screened for instrumental variables. Data from the ENIGMA consortium's GWAS meta-analysis of morphological changes in the brain were summarized as outcomes (n=15, 640). The relationship between gut microbiota and longitudinal changes in brain structure was detected using random effects inverse variance weighting (IVW), MR-Egger, MR-Weighted Median, MR-Simple Mode, and MR-Weighted Mode, and sensitivity analysis was performed. IVW estimates showed that Verrucomicrobia (β=-18.15, 95%CI: -30.25 to -6.05, P=3.27E-03) significantly attenuated the longitudinal volume changes of the putamen, Firmicutes (β=10.77, 95%CI:3.30 to 18.25), P=4.73E-03), Cyanobacteria (β=7.68, 95%CI:2.46 to 12.90, P=3.93E-03) and Gastranaerophilales (β=9.54, 95%CI:4.41 to 14.68, P=2.69E-04) significantly increased the longitudinal volume variation of pallidum. Our mendelian randomization (MR) analysis showed a causal relationship between the gut microbiota and longitudinal changes in brain structure and supported the existence of a gut-brain axis. This study provides new insights into further mechanistic and clinical studies of microbiota-mediated neuropsychiatric disorders.

Supplementary Material 1: Comparison between the performance of ALR and CLR

Little is known about the development of the tonsillar microbiota and the factors determining the establishment and proliferation of disease-associated bacteria. In this study we sampled the tonsillar microbiota of 63 piglets from 21 different litters after birth and at week 1, 3, and 5 of life. The piglets were free from disease and antimicrobial treatment throughout the study period. 16S rRNA gene V3-V4 region Illumina NovaSeq sequencing produced a minimum of 42537 reads per sample after processing with DADA2. The dataset includes 252 piglet microbiota samples and 12 samples collected from unrelated sows at the same farm. Analysis revealed that a diverse tonsillar microbiota is established shortly after birth, and that gradual change takes place during the first 5 weeks of life without weaning having a large impact. We found a strong litter effect, with siblings sharing a more similar microbiota compared to non-sibling piglets. Co-housing in rooms, within which litters were housed in separate pens, also had a large impact on microbiota composition. Sow parity and pre-partum Streptococcus suis bacterin vaccination had weaker but significant associations with microbiota composition, with piglets of vaccinated sows having a lower S. suis abundance 1 week after birth.This dataset consists of 16S rRNA gene V3-V4 amplicon sequencing of tonsillar swabs of piglets. The piglets were sampled shortly after birth, at week 1, week 3, and (after being weaned at week 4), week 5. No antimicrobial treatment was applied, and all piglets were healthy; piglets that died during the experiment were not included for sequencing. Date Submitted: 2021-11-04

Longitudinal analysis of the microbiota composition and enterotypes of pigs from post-weaning to finishing. The raw sequences of 16S rRNA genes are stored on ncbi. SubmissionID: SUB6518439 BioProject ID: PRJNA588139

Metagenomic time-course studies provide valuable insights into the dynamics of microbial systems and have become increasingly popular alongside the reduction in costs of next-generation sequencing technologies. Normalization is a common but critical preprocessing step before proceeding with downstream analysis. To the best of our knowledge, currently there is no reported method to appropriately normalize microbial time-series data. We propose TimeNorm, a novel normalization method that considers the compositional property and time dependency in time-course microbiome data. It is the first method designed for normalizing time-series data within the same time point (intra-time normalization) and across time points (bridge normalization), separately. Intra-time normalization normalizes microbial samples under the same condition based on common dominant features. Bridge normalization detects and utilizes a group of most stable features across two adjacent time points for normalization. Through comprehensive simulation studies and application to a real study, we demonstrate that TimeNorm outperforms existing normalization methods and boosts the power of downstream differential abundance analysis.

This dataset is a combination of data obtained from a longitudinal live trapping study of wild rodents in Wytham Woods, Oxford (51.796 N,-1.367 W); October 2015-18), a dissection study of rodents caught in the same woodland (October 2017-18), and a diet shift experiment on a captive colony of wood mice housed at the University of Edinburgh (May 2017). The longitudinal live trapping study dataset contains trapping data and data on the gut microbiome composition, diet and gut parasite infection of individually-identifiable rodents. Three species of rodents were trapped with Sherman live-traps fortnightly for 3 years: wood mice (Apodemus sylvaticus), yellow-necked mice (Apodemus flavicollis) and bank voles (Myodes glaerolus). Upon capture, they were injected with a passive integrated transponder (PIT) tag, measured, weighted, sexed, aged and a faecal sample was collected from individuals for microbiome, diet and parasite analyses. All rodents were released to their location of capture. The dissection study contains trapping data, gut microbiome and parasite infection data. Wood mice were trapped fortnightly for one year with Sherman live-traps at least 300m away from the longitudinal sampling grid. Individuals that had been captured and marked with a PIT tag as part of the longitudinal study were released along with other rodents species and juvenile or pregnant individuals (only non-marked adult wood mice were sampled). Wood mice were euthanized (with ethical approval) and their gastrointestinal tract dissected for counts of gut helminths. Samples from along the gastrointestinal tract were taken for gut microbiome analysis. The diet shift experiment dataset contains data on experimental diet treatments and gut microbiome composition of wood mice captively bred in a facility at the University of Edinburgh. Wood mice were given diets varying in the ratio of food supplementation (dried mealworm and/or peanut) and faecal samples taken periodically over 30 days to measure changes in gut microbiome composition and function. Samples for microbiome and diet characterization were stored without buffer at -80̊C. Samples for parasite detection were stored in 10% formalin and refrigerated at 4̊C. This work was funded by a NERC independent Research Fellowship. Full details about this dataset can be found at https://doi.org/10.5285/4f454849-0d30-4c27-ad5b-d285e461bedc

Wildlife microbiome studies are being used to assess microbial links with animal health and habitat. The gold standard of sampling microbiomes directly from captured animals is ideal for limiting potential abiotic influences on microbiome composition, yet fails to leverage the many benefits of non-invasive sampling. Application of microbiome-based monitoring for rare, endangered, or elusive species creates a need to non-invasively collect scat samples shed into the environment. Since controlling sample age is not always possible, the potential influence of time-associated abiotic factors was assessed. To accomplish this, we analyzed partial 16S rRNA genes of fecal metagenomic DNA sampled non-invasively from Rocky Mountain elk (Cervus canadensis) near Yellowstone National Park. We sampled pellet piles from four different elk, then aged them in a natural forest plot for 1, 3, 7, and 14 days, with triplicate samples at each time point (i.e., a blocked, repeat measures (longitudinal) study design). We compared microbiomes of each elk through time with point estimates of diversity, bootstrapped hierarchical clustering of samples, and a version of ANOVA–simultaneous components analysis (ASCA) with PCA (LiMM-PCA) to assess the variance contributions of time, individual and sample replication. Our results showed community stability through days 0, 1, 3 and 7, with a modest but detectable change in abundance in only 2 genera (Bacteroides and Sporobacter) at day 14. The total variance explained by time in our LiMM-PCA model across the entire 2-week period was not statistically significant (p>0.195) and the overall effect size was small (<10% variance) compared to the variance explained by the individual animal (p<0.0005; 21% var.). We conclude that non-invasive sampling of elk scat collected within one week during winter/early spring provides a reliable approach to characterize microbiome composition in a 16S rDNA survey and that sampled individuals can be directly compared across unknown time points with minimal bias. Further, point estimates of microbiome diversity were not mechanistically affected by sample age. Our assessment of samples using bootstrap hierarchical clustering produced clustering by animal (branches) but not by sample age (nodes). These results support greater use of non-invasive microbiome sampling to assess ecological patterns in animal systems. Methods 1.1 Sample collection Scat samples from 4 elk were collected near the northern boundary of Yellowstone National Park in Montana in March 2016. Animal sampling was conducted non-invasively within 15 minutes of defecation. Elk sex and age could not be accurately determined due to these samples being collected after observing the elk defecating from a distance using binoculars. Based on our observations, they were most likely adult females or young males. Fecal samples from each scat pile (i.e., individual) were collected from the ground with sterile gloves and forceps and placed in sterile whirl-pak sample bags. Sample whirl-paks were placed on wet ice in a cooler in the field for transportation to the experimental site. The experimental site was located on a sparsely forested plot near Evaro, MT with conditions known to be suitable as elk habitat, at approximately 4000 ft elevation. Three pellets from each animal were frozen at -20° C after arriving at the experimental site approximately 6 hours post-defecation. This initial subsample represents time-point zero samples (and technical replicates) with minimal exposure to ambient conditions typical of a direct or capture-based sampling scheme. The remaining pellets from each elk were placed in square plastic culture plates (25 cm x 25 cm) with a grid backing using sterile gloves and forceps. Each culture plate had a larger glass plate suspended above it at a height of 4 cm using a cork stopper in each corner to allow air flow and prevent direct contact with incidental precipitation (although no precipitation occurred on-site during the study), and the group of culture plates was surrounded by protective wire fencing. One plate was used for each technical replicate, with each replicate plate containing samples from all four individuals (for photos of the enclosure and a schematic of the experimental layout see Supplemental Figure 1). The samples were exposed to ambient conditions from March 27th through April 9th (14 days). Three samples from each elk were removed from the replicate plates after 1 day, 3 days, 7 days, and 14 days and immediately frozen at ‑20° C after removal from ambient conditions. A total of 60 elk pellets were experimentally collected. Temperature was logged in 10-minute increments during the study using Thermocron temperature loggers (OnSolution Pty Ltd, Australia) distributed above and below the culture plates and shielded from direct sunlight. The temperature data were aggregated into hourly oscillations, daily max and minimum, and a smoothed average temperature. Additional temperature recordings were obtained from a NOAA weather station (Point 6, MT) 3.5 miles and 4000 ft above our site as reference. 1.1 Sample preparation, DNA extraction and sequencing Frozen elk fecal pellets (stored frozen at -20° C) were prepared for DNA extractions by separating a standard weight (250 mg) cross-section from each pellet using a sterile petri dish (10 cm) and sterile safety razor blade for each sample. This fraction was placed into a designated sample tube from the Qiagen PowerSoil DNA extraction kit (Qiagen Inc., Germantown, MD) and processed using the manufacturer’s recommended protocol. The resulting purified metagenomic DNA was eluted with 100 µL PCR-grade water and stored at -20° C prior to further analysis. To assess the bacterial community present in the fecal DNA extraction, we used a generally-conserved (i.e., “universal”) 16S/18S barcoded primer set (536F and 907R) designed to amplify the V4 and V5 variable regions of the rRNA gene (Holben et al., 2004) and PCR using 1mL of elk fecal sample metagenomic DNA standardized to 25ng/mL as template. Once amplified, samples were gel purified using the QIAGEN Gel Purification kit (QIAGEN, Germantown, MD) following the manufacturer’s recommended protocol for downstream direct sequencing. An Illumina MiSeq platform (San Diego, CA, USA) was used to obtain 300 base-pair (bp) paired-end sequencing using the Illumina MiSeq Reagent Kit.