Is unrestricted taxonomic assignment better for estimating numbers of genera?

[Jan_Kollar (Jan Kollár)]

Hi @SoilRotifer,

@SoilRotifer

This made me realize that I should have suggested that you try using the weighted classifiers, these are also available on the Data resources page...

Very interesting suggestion and an excellent explanation, will definitely try that, thank you!

@SoilRotifer

Perhaps you can align these amplicons to some reference genera and see how they phylogenetically cluster?

Did that, helped with resolving some ASVs at genus level (and, interestingly, the unrestricted classification was correct there at the genus level as well! Plus BLAST also suggested the same as the top hit, so yay!).

@SoilRotifer

You can also use RESCRIPt to curate your own reference database. For example, you might consider some of the other options for the qiime rescript dereplicate ... command compared to what is outlined in the tutorial.

Actually, I used your excellent tutorial while building my custom 18S diatom reference database, along with this one. Extremely useful, thank you! I'll check also the one you suggest.

@SoilRotifer

This sounds like a really cool project! Reminds me of my micro-invertebrate days.

Yeah, it's very small, very simple (relatively), and a lot of fun. With 113 ASVs, it's possible to basically pay attention to every one of them, which is kind of cool, especially compared to working with bacteria. Soon, I'll have to move to our global dataset with several hundreds of samples and I have some feeling that I'll look back to these humble beginnings with tears of pain in my eyes

Anyway, thank you very much for your time and ideas, it's a great help!