How to interpret the Taxa assignment results!

General Discussion
1

I have some questions about the results of taxa assignment. I know not all ASV can be assigned to the exact species level.
For example, that I have these two results:

1>d__Bacteria;p__Firmicutes;c__Bacilli;o__Lactobacillales;f__Enterococcaceae;g__Enterococcus;__

2>d__Bacteria;p__Bacteroidota;c__Bacteroidia;o__Bacteroidales;f__Barnesiellaceae;;

I am confused about the 1> and 2>

For 1>, there is only 1 ASV right?, but we don’t know a exact the name of species?

For 2>, I am not sure how many ASV? Is it only one ASV? Or a group of ASVs can only be assigned to unknown genus level…

Thanks,

2

Hi!

It can be 1 ASV, but also it can be (and most probably) a bunch of them. Annotation of amplicon sequencing is working better on genera level, than on species, due to the lack of variability in targeted rRNA regions. So, in some cases classifier can not assign with enough confidence level species, genera, families and so on.

The same as above. Most probably a bunch of ASVs, that can not be assigned to certain taxonomy level with enough confidence level.

1 Like
3

Thank you.

1> If more than one ASV could be assigned to unknown species or genus, can I find which ASVs assgin to this unknown genus or species? Can I extract the related reads? If so, which scripts should I use

2> In another case, if the ASV has assigned to an exact species, for example, E.coli, does this mean only one ASV is identified? Or it could be more than one ASV identified as E. coli?
Thanks

4

One way is to visualize your taxonomy.qza file (convert to qzv) and search for ASV ids that were assigned to taxa of interest. Or you can export taxonomy.qza as .tsv file and perform search in excel or another software. If you also want to obtain sequences itself, you can visualize rep-seqs.qza and search by ASV ids, or export rep-seq.qza as fasta file and search there by ASV ids. I am not aware if there is some tricky way to perform it by Qiime2 plugins, but you can do it manually for a small amount of taxa/ASVs. If you want to check a bunch of them, you will need some scripting in any suitable language.

Species are not equal to ASVs. In some cases, it can be, that only one ASV was assigned to one of the species. But in most cases, you will find that several ASVs are assigned to one species.

One ASV can be assigned only to one taxa. But one taxa may include numerous ASVs.

7

Hi @sdpapet
Looks like I missed one option that can help you.
Here is the link to the solution that may work for you:
https://docs.qiime2.org/2021.4/tutorials/metadata/#exploring-feature-metadata

Thank you @Nicholas_Bokulich for your help

1 Like