Sixteen samples (four groups of four samples) were collected and analyzed by 454 Flx pyrosequencing, and comparisons were made between Herd 1 and Herd 2, between Herd 1 Time 1 and Herd 1 Time 2 tissues, and between tissue and

brush samples from Herd 1 Time 2 in these results. Bar-coded 16S pyrosequencing A total of 210,433 quality reads were obtained from the four groups of pigs sampled, with at least 15,000 reads per group. Samples of tonsil tissue from Herd 1 at time 2 yielded the fewest number of quality reads. Table 1 shows the number CP-868596 supplier of reads obtained from each of the four groups of pigs and the percent of those reads that could be taxonomically assigned at a 60% confidence level using the RDP Classifier. Overall, greater than 97% of the total reads could be taxonomically assigned at the phylum, class, and order level. This dropped to 90.5% at the family level and further dropped to 72.3% at the genus level. Taxonomic assignment of reads was consistently learn more lower at all levels for Herd 2 compared to all three

groups of samples from Herd 1. Table 1 Taxonomic characterization of tonsillar microbial communities   Sample # Readsa Phylumb Classb Orderb Familyb Genusb Herd 2 Tissue 99894 95.6% 95.4% 94.8% 82.7% 64.7% Herd 1 Time 1 Tissue 54932 99.7% 99.6% 99.1% 96.7% 85.0% Herd 1 Time 2 Tissue 15929 99.8% 99.5% 99.4% 98.7% 70.1% Herd 1 Time 2 Brush 39678 99.9% 99.5% 99.5% 98.6% 75.0% Total # reads   210433 205795 205346 204467 190540 152192 Avg % Assigned     97.8% 97.6% 97.2% 90.5% 72.3% a the sum of all sequences of 4 individuals b%

of reads taxonomically assigned at each level Figure 1 shows the rarefaction plots for Suplatast tosilate the four groups. Herd 1 and Herd 2 plots demonstrate that Herd 2 had significantly more phylotypes and greater unsampled diversity (Figure 1A). Comparison of the three groups of Herd 1 pigs reveals similar trajectories even though the number of reads sampled varied (Figure 1B). Taken together, this suggests that the microbial community in the tonsils in Herd 2 was more complex at this level of interrogation. Figure 1 Rarefaction curves computed with the RDP Pyrosequencing Pipeline. Rarefaction curves are presented for each group of samples obtained by 454 pyrosequencing. The curves for herds 1 and 2 at time 1 are shown in panel A, while the curves for all three groups of samples from herd 1 are shown in panel B. As stated above, a total of 210,433 reads was obtained for the four groups. Table 2 indicates the number of reads made from each individual sample as well as the total for each group. The number of reads for each individual and each group forms the basis of the comparisons for the number of OTUs, Chao-1 richness, and the Simpson diversity indices. Using a cutoff of 97% identity for species level distinctions, the number of OTUs detected per sample ranged from 57 to 730.