thermophilus for SGII (Panel B) and SGI (Panel C) spacers. In panels B and C, each box represents a spacer in a CRISPR locus in the CRISPR Database, and colored boxes represent NVP-BGJ398 cost spacers that also were present in this study. White boxes represent spacers that were not identified in this study. In each subpanel, the colored LY2874455 manufacturer boxes from the top locus represent spacers that were matched by skin-derived spacers, and the bottom box represents spacers that were matched by saliva-derived spacers. To determine whether skin-derived CRISPR spacers

matched viruses present in the saliva, we sequenced the viromes present in each of our subjects’ saliva Selleckchem Geneticin on Day 1 and Week 8. Similar to our previous studies [14], the proportion of CRISPR spacers matching

virome reads was relatively low. When examining the pooled reads from all subjects, we found that between 0.0% and 1.0% of the CRISPR spacers in each subject matched virome reads for SGI spacers and SGII spacers (Additional file 2: Figure S7). When we tested the skin- and saliva-derived spacers against a larger database of salivary viromes from a cohort 21 human subjects [10], we found that a high number of salivary- and skin-derived spacers matched salivary virome reads (range from 14 to 60% for SGII spacers and 10 to 24% for SGI spacers). The proportion PDK4 of spacers matching salivary viruses was significantly (p ≤ 0.002) higher for saliva-derived spacers than

for skin-derived spacers for Subjects #3 and #4 for SGII spacers, but not Subjects #1 and #2. There also were a significantly higher proportion of SGI saliva-derived spacers that matched salivary viruses in Subjects #2 and #3, but not Subjects #1 and #4 (Figure 8). Figure 8 Percentage of SGI (Panel A) and SGII (Panel B) CRISPR spacers matching virome reads from the saliva of 21 human subjects [10]. The Y-axis shows the mean percentage of the CRISPR spacers from all time points combined that matched virome reads from the cohort of 21 subjects. The X-axis represents the saliva- and skin-derived spacers for each subject. Standard error bars are represented above each bar, and the p-value is demonstrated above each error bar. Subjects 1 through 4 are shown consecutively from left to right on the X-axis. We also tested whether there were matches to spacers found in previously sequences metagenomes recovered from the human oral cavity [39], the gastrointestinal tract [40], and human skin [41]. We found that a significantly higher percentage of SGII (3-4%) and SGI (4-5%) spacer sequences were found in oral metagenomes than the 1-2% of SGII and SGI found in the gut and the <1% found on the skin (p < 0.02) (Additional file 2: Figure S8, Panels A and B).