Table 3.
Bioinformatic parameters and databases used in the analysis of bar-coded pyrosequencing results
| Step | Parameter | Targeted rDNA Gene | |
|---|---|---|---|
| 16S | 18S | ||
| Preprocessing | Length threshold | 370 | 300 |
| Number of ambiguities tolerated | 0 | 0 | |
| Detection of proximal primer sequence | Complete and perfect | Complete and perfect | |
| Detection of distal primer sequence | No | Perfect, but potentially incomplete | |
| Clustering | Chosen level of similarity (%) | 95 | 95 |
| Ignoring differences in homopolymer lengths | Yes | Yes | |
| Filtering | Chosen clustering similarity threshold | 95 | 95 |
| Used taxonomic database | SILVA (r114) | SILVA (r111) | |
| Chosen taxonomic level | Phylum | Phylum | |
| Similarity or confidence threshold (%) | 90 | 85 | |
| Homogenization | High-quality reads kept for each sample | 2322 | 4378 |
| Taxonomy | Used taxonomic database | SILVA (r114) | SILVA (r111) |
| Method or tool of comparison | usearch | megablast | |
| Similarity or confidence threshold (%) | 80 | 80 | |
| Analysis | Chosen level of similarity (%) | 95 | 95 |
| Ignoring differences in homopolymer lengths | Yes | Yes | |
| Computation of a unifrac distance matrix | Yes | Yes | |