SEARCH

SEARCH BY CITATION

Fig. S1. Rarefaction analysis of SGII CRISPR spacers in the saliva of human subjects at each sampled time point (A–D) and all time points combined (E). Rarefaction curves were created using 10 000 random iterations based on spacer richness.

Fig. S2. Heatmap of unique SGII spacers present in each subject at all time points. Each row represents a unique spacer sequence, and the columns represent each subject atdifferent time points. The intensity scale bar is located to the right, and its values correspond to the percentage of spacers present in each subject and time point.

Fig. S3. Diagram of SGI CRISPR spacer homologous to various bacteria, phage and plasmids. For the initial time point for each subject, homologues to CRISPR spacer complements are demonstrated. At each subsequent time point, only homologues to newly identified spacers that were not present in prior time points are shown. For example, in subject #3, spacers homologous to streptococcal phage PH-10 are identified on Day 1, while on Day 30 newly identified spacers that were not present on Day 1 also have homology to phage PH-10. (A) Subject #1, (B) Subject #2, (C) Subject #3 and (D) Subject #4. Homologues to phage are shown in red, plasmids in blue, and bacteria in green.

Fig. S4. Analysis of shared SGII CRISPR spacers between different human subjects.

Fig. S5. Principal coordinates analysis of SGII CRISPR spacer composition from human saliva based on beta diversity. Green – subject #1, gold – subject #2, magenta – subject #3 and cyan – subject #4. M-6 represents Month -6, M-3 represents Month -3, D1 represents Day 1, D30 represents Day 30, D60 represents Day 60 and M11 represents Month 11.

Fig. S6. Virome reads with matches to both SGI and SGII CRISPR spacers. The percentage each virome that has unique reads with CRISPR spacer matches is demonstrated on the y-axis. Matches to the salivary viromes in this study are represented in blue, matches to human oropharyngeal viromes are represented in red, matches to human respiratory tract viromes are represented in green, matches to human fecal viromes are represented in orange, and matches to a marine virome is represented in yellow.

Fig. S7. Virome reads with matches to either SGI (A) or SGII (B) CRISPR spacers. The number of unique reads with spacer matches in each virome from each subject is demonstrated on the y-axis. Unique virome reads with spacer matches is shown in green for subject #1, gold for subject #2, magenta for subject #3 and cyan for subject #4.

Fig. S8. Percentage of spacers with matches to virome reads that are exact or have single nucleotide mismatches. A. SGI CRISPR spacer-read matches. B. SGII CRISPR spacer-read matches.

Fig. S9. Fraction of virome reads with SGII CRISPR spacer matches over time within each subject. The virome reads with CRISPR spacer matches are normalized by virome size. Each column represents the SGII CRISPR spacer repertoire characterized at the individual labelled time point, and the y-axis represents the normalized percentages of unique reads with matches to CRISPR spacers recovered from each of the time points. Blue represents spacer-read matches from Month -6, red represents Month -3, green represents Day 1, purple represents Day 30, cyan represents Day 60 and orange represents Month 11.

Fig. S10. CRISPR spacer-virome read matches for each subject and each time point. The fraction of virome reads with corresponding CRISPR spacer matches for each time point is represented for each subject. The y-axis represents the fraction of virome reads from each of the subjects (colour-coded) with corresponding spacer matches, and the x-axis represents the time point from which the CRISPR spacers were analysed. Subject #1 spacers are represented in red, subject #2 spacers are represented in orange, subject #3 spacers are represented in green, and subject #4 spacers are represented in blue. Panels on the left represents SGI spacer comparisons, and panels on the right represents SGII spacer comparisons.

Fig. S11. Heatmap of virome contigs with SGI and SGII CRISPR spacer matches. Each row represents a unique virome contig, and each column represents the set of CRISPR spacers found on that day for that subject. Virome contigs from each subject are identified on the left of each diagram, and the day is shown on the right of each diagram. The intensity scale bar is shown below each panel, and its values correspond to the percentage of CRISPR spacer matches present in each subject and time point. A. SGI CRISPR spacer-read matches. B. SGII CRISPR spacer-read matches.

Fig. S12. Virome contig from subject #2 at Month -6. Putative open reading frame prediction was performed using FGenesV, and putative identification based on the presence of homologues using tblastX analysis (E-score cut-off < 10−5) of the NCBI non-redundant database. Open reading frames with CRISPR spacer matches are demonstrated in red, and those with spacer matches are labelled with the number of unique spacers that match. A distance scale bar is shown below the diagram (9873 nucleotides, average coverage 8×).

Fig. S13. Analysis of PAMs for spacer-read matches for SGI and SGII. The diagram shows the 5′ eight nucleotides (A and C) and the 3′ eight nucleotides (B and D) located within the virome reads that have spacer matches. The x-axis represents the position relative to the CRISPR-spacer match, and the y-axis shows the relative conservation of nucleotides at each site measured in bits. (A and B) SGI, (C and D) SGII.

Table S1. Subject demographics.

Table S2. Virome read characteristics.

Table S3. SGII CRISPR repeats for all subjects.

Table S4. SGII spacer hits.

FilenameFormatSizeDescription
EMI_2775_sm_Fig-LegendsS1-13.doc45KSupporting info item
EMI_2775_sm_FigS1-13.pdf2517KSupporting info item
EMI_2775_sm_TabS1-4.doc133KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.