The mean percent alignments of the individuals were used in Figure  GW3965 clinical trial 4 and Additional files 4 and 5. The normalized mean percent of ORFs in each functional category was used in Figures  5 and 6. Metagenome

comparisons were statistically compared by Student’s t-tests (P < 0.05) using SigmaPlot (Systat Software, Inc., San Jose, CA, USA). Immune-modulatory motif identification An identity of 100% was used to search for immune-modulatory motifs by alignment with assembled contigs from the human milk metagenome (56,712 contigs) or the fecal metagenomes described above (834,774, 64,662 and 553,391 contigs from BF-infants’, FF-infants’ and mothers’ feces, respectively). The human genome (2,865,822,365 bp) was used as a

comparative reference. Z-score was calculated using the formula Z = (O-E)/Stdev, where O was the observed number of hits and E was the expected number of hits using the formula E = (L cont )(N h/L h ) where L cont was length of sequences or assembled contigs, N h was number of sites found in the human genome (or compiled bacterial genomes); Stdev was Selleckchem QNZ the standard deviation of occurrence of each motif in 22 + X + Y human chromosomes. Availability of supporting data The data set supporting the results of this article is available in the MG-RAST repository, under the project name Human_milk_microbiome, http://​metagenomics.​anl.​gov/​linkin.​cgi?​project=​2959. Acknowledgements This work was funded by the 2-hydroxyphytanoyl-CoA lyase Canadian Institutes of Health Research, Institute of Nutrition, Metabolism and Diabetes (grant 82826 to IA) and Canada Foundation for Innovation, Leaders Opportunity Fund/Ontario Research Fund (grant 22880 to II). TLW is HDAC inhibitor review supported by a Natural Sciences and Engineering Research Council (NSERC) Canadian Graduate Scholarship. We are grateful to Lynne Cullen and Dr. JoAnn Harrold of the Children’s Hospital of Eastern Ontario for donor recruitment and milk collection. We would

also like to thank Dr. Will Spencer of BMI for isolating DNA from human milk, Kathy Sheikheleslamy of StemCore Laboratories (Ottawa Hospital Research Institute, Ottawa, Canada) for her sequencing efforts, and Chris Porter and Gareth Palidwor for filtering Illumina outputs. Electronic supplementary material Additional file 1: Abundance of DNA fragments in pooled human milk, sequenced seven times. This table contains the number of DNA sequences per run and their general alignments. (DOCX 12 KB) Additional file 2: Classification of 51 bp DNA sequences derived from human milk by best hit analysis. This table contains all genera with at least one alignment match to sequences from human milk-derived DNA. (DOCX 22 KB) Additional file 3: Predicted open reading frames from human milk DNA sequences aligning to rRNA genes of known organisms.