Hi @jrhaulung,
It appears the SILVA does not have many representatives for Paramecia. If you go here and search for Paramecium bursaria you'll see that there are ~ 29 entries. But if you search for Paramecium you'll observe 262 references. Many of them quite short and may not span the V1V2 region.
I downloaded your data and trained the calssifier on the full SILVA 138.2 database, and essentially observed the same results. Given the lack of references I mentioned above and the many hits I obtain with GenBank, perhaps make your own database from GenBank? Many of your reads hit quite well there.
- 16S only.
- Use taxonomy IDs (using CO1 as an example).
- A modified approach using taxonomy IDs (using trnL as an example).
You can modify the above tutorials as needed to pull down the taxonomic groups / gene sequences you are interested in.
-Mike