Do you need to BLAST sequences classified with a SILVA classifier?

[Ellenphant (Ellen)]

I have paired-end 16S sequences that I have taxonomically classified with a SILVA database classifier.
Is there any benefit to using NCBI to blast the sequences as well? I guess it could potentially provide more phylogenetic information but couldn’t it also provide hits for anything that matches?

[Nicholas_Bokulich]

Hi @Ellenphant,

@Ellenphant

I have paired-end 16S sequences that I have taxonomically classified with a SILVA database classifier.

A naive Bayes classifier trained on the SILVA database using q2-feature-classifier? Great! As long as the results look good, that should be all you need.

@Ellenphant

Is there any benefit to using NCBI to blast the sequences as well?

Only if you need a second opinion or want to figure out what some unclassified reads are. NCBI BLAST can be quite misleading for taxonomic classification as I've described here:

@Nicholas_Bokulich

NCBI BLAST results are misleading — of course you get species-level classifications, because you are performing local alignment against other (usually longer) sequences that typically have species annotations. Just because a short read aligns to a reference, even perfectly, does not mean that is a correct match.

It is important to assess (1) how good is that match, e.g., how much coverage and mismatches and (2) how many other taxa have equally or similarly good hits?

Assessing the quality of matches can be an arduous process, especially if you have 100s or 1000s of sequences — and this is why other methods are used by QIIME 2 (and similar platforms) for taxonomic classification: to automate the process of taxonomic classification.

All that said, seeing unclassified sequences classify to bacteria with NCBI BLAST can sometimes indicate an issue with the database/classifier you are using, or with the query sequences.

@Ellenphant

I guess it could potentially provide more phylogenetic information but couldn’t it also provide hits for anything that matches?

Exactly! So you need to look carefully to see how similar those hits are, and whether you have multiple ties for top hit. It gets messy, which is why we have methods in q2-feature-classifier to do the job for you.

I hope that helps!