Classification of metagenomic fragments was undertaken KU55933 mw using the Pplacer package v1.1 alpha11 [16]. The taxonomic assignment of each reference sequence was retrieved from the NCBI taxonomy database using

Taxtastic (http://​fhcrc.​github.​com/​taxtastic) and a Pplacer reference package was created for each KO of interest. Metagenomic sequence fragments were then placed on the tree using Pplacer. This allowed for assignment of each ORF to a taxonomic attribution with a high level of confidence. These classifications were then retrieved using the guppy classification method of Pplacer, which reports the closest taxonomic attribution for each phylogenetically placed read. Differences in abundances of species between lean and obese patients were examined using STAMP version 2 employing the Welch two-sided t-test with Bonferroni multiple test correction and a 0.05 p-value cut-off. Acknowledgements We would like to thank Donovan Parks, Robert Eveleigh, Morgan Langille and Erick Matsen for assistance with statistical analysis, alignment processing, phylogenetic clustering and taxonomic assignments. This work is supported by CIHR grant number CMF-108026. RGB acknowledges the support

of Genome Atlantic and the Canada Research Chairs program. Electronic supplementary material Additional file 1: Figure S1. Phylogenetic trees of K02031-K02035 (A-E respectively) showing the spread of gut-associated species. Phylogenetic analysis of each set of sequences from proteins within the peptides/nickel transporter showing the spread of gut-associated species (red terminal branches) throughout each Regorafenib molecular weight tree. (PDF 523 KB) Additional file 2: Table S1. Consistency index between KO trees of gut-associated species and taxonomic ranks. Subtrees for each KO comprising only gut-associated species were examined for consistency between taxonomy and phylogenetic placement. (PDF 16 KB) Additional file 3: Figure S2. Phylogenetic tree of gut-associated species for K02031. Phylogenetic analysis of

only gut-associated species showing the spread of Faecalibacterium prausnitzii (green) and Clostridium difficile (red) strains. (PDF 31 KB) Additional file 4: Figure S3. Phylogenetic analysis of proteins associated with K02031-K02035 within Faecalibacterium prausnitzii. Resminostat Protein sequences annotated as being part of the nickel/peptides transporter complex (K02031-K02035) within the five strains of F. prausnitzii were found to fall into one of six subtrees within each protein tree. Each subtree corresponds to an operon as listed in Figure 2. IMG gene object ID locus names for sequences are listed beside the strain name. Branch labels correspond to bootstrap values. Branch lengths are not to scale. (PDF 226 kb) (PDF 227 KB) References 1. Bäckhed F, Ley RE, Sonnenburg JL, PF299804 chemical structure Peterson DA, Gordon JI: Host-bacterial mutualism in the human intestine. Science 2005, 307:1915–1920.PubMedCrossRef 2.