Naive Bayes classifier for the 515F/806R region


I used the Naive Bayes classifier trained on primers 515F/806R available on Data resources — QIIME 2 2022.2.0 documentation. I recently notice that in the earth microbiome protocol their are two versions of primer for the 515F/806R regions.

Updated sequences: 515F (Parada)–806R (Apprill), forward-barcoded:
Original sequences: 515F (Caporaso)–806R (Caporaso), reverse-barcoded:

After a thorough search on qiime forum and internet I did not find information regarding which primer set was used to train the classifier. Had you inform if the this classifier used the updated or original sequences?\

Best regards,

Hi @Antonio_Louvado, welcome to :qiime2:!

You can determine this by looking through the provenance information of the silva-138-99-515-806-nb-classifier.qza file. That is, simply drag-drop this file into QIIME 2 View, then click on the provenance tab. From here you can click on each item in the provenance graph to see what parameters were used for each command. Like this:

Here you can see, the primer sequence that where used. It looks like the original primer sequences where used.

Personal note: The only difference between these two primer sets is the additional IUPAC ambiguity bases in the "updated" primer sequences. When I make the classifier I use the "updated" primer set, as they are more "universal" than the original sequences. Which means better success in finding and searching for your amplicons.

1 Like

Thanks very much for the information.


This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.