Hi,
I have similar issue such that a large portion of my taxa are unassigned or bacteria only with no info about other levels. When I blast the sequences in NCBI, I get macaque (which is one of my samples) and for many others nothing even though QIIME was able to find some. I am dealing with blood so biomass is very low. What can I do to resolve this issue?
This is a pretty good confirmation that you have mostly (or entirely) host DNA in your samples. Most of the QIIME 2 classifications are to mitochondrial DNA, which is clearly host. So that corroborates NCBI BLAST's match to macaque (the SILVA database you are using does not have that much detail on where the mitochondrial reads are coming from!)
The many other unassigned reads are most likely other segments of host DNA — they clearly do not match to any bacterial reads. They may also be sequence noise and reagent contamination, which are both big problems with low-biomass samples.
I'm not so sure QIIME found any real bacterial reads in your samples. The vast majority of reads are either "unclassified" (probably host DNA) or "mitochondria" (there are 2 different taxonomic groupings labeled as mito DNA). The remainder are mostly underclassified but belong to the same lineage as mitochondria, so are probably other mitochondrial reads that are not being fully classified, maybe because they contain noise or are shorter reads. This includes even the large group that is only labeled D_0__Bacteria — it's all mito DNA.
The only non-mito is a couple groups labeled as Archaea. Unless if you are expecting these in your samples, I would venture a guess that this could be reagent contamination (which, again, is a very big issue with low-biomass samples).
Sorry to say, but based on the fact that these are low-biomass samples I am not sure there is anything being detected other than host DNA 
So I do not think this is a problem with the classifier or the parameters you are using — you may need to select a different primer set that does not amplify mito DNA.
oh that is a bummer 
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.