We present QIIME 2, an open-source microbiome data science platform accessible to users spanning the microbiome research ecosystem, from scientists and engineers to clinicians and policy makers. QIIME 2 provides new features that will drive the next generation of microbiome research. These include interactive spatial and temporal analysis and visualization tools, support for metabolomics and shotgun metagenomics analysis, and automated data provenance tracking to ensure reproducible, transparent microbiome data science.

In the version of this article initially published, some reference citations were incorrect. The three references to Jupyter Notebooks should have cited Kluyver et al. instead of Gonzalez et al. The reference to Qiita should have cited Gonzalez et al. instead of Schloss et al. The reference to mothur should have cited Schloss et al. instead of McMurdie & Holmes. The reference to phyloseq should have cited McMurdie & Holmes instead of Huber et al. The reference to Bioconductor should have cited Huber et al. instead of Franzosa et al. And the reference to the biobakery suite should have cited Franzosa et al. instead of Kluyver et al. The errors have been corrected in the HTML and PDF versions of the article.

36Background: Taxonomic classification of marker-gene (i.e., amplicon) sequences 37

Background

Background

BackgroundThe gut microbiota, the aggregates of microbial cells that inhabit the gastrointestinal tract, communicates bidirectionally with the brain via immune, neural, metabolic, and endocrine pathways, known as the gut‐brain axis. The gut‐brain axis is suspected to contribute to the development of Alzheimer’s disease (AD). We hypothesize that altered gut microbiota composition contributes to the development of AD pathologies and neuroinflammation via the gut‐brain axis.MethodTo characterize the gut microbiota of 3xTg‐AD mice modeling plaque deposition and hyperphosphorylated tau, fecal samples were collected fortnightly from 4 to 52 weeks of age (n=57 3xTg‐AD mice, n=31 wild‐type). The V4 region of the 16S rRNA gene was amplified and sequenced on the Illumina MiSeq. Data were analyzed using QIIME 2. Targeted reverse transcription qPCR assays were used to assess inflammation in the hippocampus and colon at 8, 24, and 52 weeks of age. Fold change was calculated using ΔΔCt.ResultOur results show altered microbial communities in 3xTg‐AD mice when compared to wild‐type [(PERMANOVA (8 weeks, p=0.001), (24 weeks, p=0.039), (52 weeks, p=0.058)]. Using q2‐longitudinal, we identified a temporal increase in Bacteroides acidifaciens and Turicibacter spp. in 3xTg‐AD mice (r‐squared = 0.658615). Using Random Forest, we successfully predicted strain in 3xTg‐AD mice 100% of the time, and in WT mice 92.85% of the time, improving accuracy over baseline assignment by 1.3 fold. Colonic expression of GFAP was increased at 24 week 3xTg‐AD mice compared to 52 week 3xTg‐AD mice (p=0.009, Mann‐Whitney). Colonic gene expression of IL‐6 was increased in 52 week 3xTg‐AD mice compared to 52 week WT mice (p=0.015, Mann‐Whitney). Hippocampal expression of GFAP was increased at 52 week 3xTg‐AD mice compared to 52 week WT (p=0.049, Mann‐Whitney). Finally, hippocampal expression of Mrc1 was elevated at 24 weeks in 3xTg‐AD mice compared to 52 weeks (p= 0.004, Mann‐Whitney).ConclusionWe have identified changes in the gut microbiota and immune response that may be predictive of the development of AD pathologies. Future shallow shotgun metagenomics sequencing will assess strain‐level features and functions of the gut microbiota in AD.

BackgroundThe gut microbiota, the aggregates of all microbial cells that inhabit the gut, bidirectionally communicates with the brain through cytokines, hormones, metabolites, and neurotransmitters via the gut microbiota‐brain axis. The gut microbiota is thought to contribute to the development of Alzheimer’s disease (AD), which is characterized by plaque deposition, neurofibrillary tangles, and neuroinflammation. We hypothesize that manipulation of the gut microbiota will alter development of AD pathologies and neuroinflammation via the gut microbiota‐brain axis.MethodWe performed fecal microbiota transplants (FMT) from aged (52‐64 weeks) 3xTg‐AD mice, which are modeling plaques and neurofibrillary tangles, to young 3xTg‐AD (n=5) or wild‐type mice (n=10) via oral gavage. Phosphate buffered saline (PBS) was gavaged into 3xTg‐AD (n=5) and wild‐type mice (n=10) as a control. At 8 weeks, mice were gavaged with FMT or PBS for 5 consecutive days, followed by fortnightly maintenance transplants for 24 weeks. The V4 region of the 16S rRNA gene was sequenced on the Illumina MiSeq. Data were analyzed using QIIME 2. Reverse transcriptase qPCR was used to assess microgliosis, astrocytosis, and Th1/Th2 inflammation in the hippocampus of the FMT cohort at 24 weeks of age.ResultWe observed a shift in microbiota composition of FMT‐treated mice when compared to control (PBS‐treated) mice. Bacteroides acidifaciens was increased in 3xTg‐AD and wild‐type mice receiving FMT. We demonstrate partial engraftment of the gut microbiota from aged 3xTg‐AD mice in all FMT‐treated mice, demonstrated by a Random Forest model, which correctly predicts treatment groups based on gut microbiota composition (Accuracy Ratio over baseline assignment: 2.6).ConclusionWe demonstrate the ability to transplant an aged gut microbiome into young mice. Future shallow shotgun metagenomic sequencing will be used to determine the species‐ and strains‐ that engraft in the GI tract. Additionally, targeted reverse transcription RT‐qPCR and immunostaining for plaques and hyperphosphorylated tau in the hippocampus will be used to assess how FMTs alter AD pathologies. These studies will contribute to our understanding of how features of the gut microbiota may contribute to AD development.

29 30 ghost-tree is a bioinformatics tool that integrates sequence data from two genetic markers 31 into a single phylogenetic tree that can be used for diversity analyses. Our approach uses 32 one genetic marker whose sequences can be aligned across organisms spanning divergent 33 taxonomic groups (e.g., fungal families) as a "foundation" phylogeny. A second, more 34 rapidly evolving genetic marker is then used to build "extension" phylogenies for more 35 closely related organisms (e.g., fungal species or strains) that are then grafted on to the 36 foundation tree by mapping taxonomic names. We apply ghost-tree to graft fungal 37 extension phylogenies derived from ITS sequences onto a foundation phylogeny derived 38 from fungal 18S sequences. The result is a phylogenetic tree, compatible with the 39 commonly used UNITE fungal database, that supports phylogenetic diversity analysis 40 (e.g., UniFrac) of fungal communities profiled using ITS markers. 41 42Availability: ghost-tree is pip-installable. All source code, documentation, and test code 43are available under the BSD license at https://github.com/JTFouquier/ghost-tree. 45PeerJ PrePrints | https://dx.doi.org/10.7287/peerj.preprints.1106v1 | CC-BY 4.0 Open Access | rec