Hi @Melissa_Hayes
First, someone with experience in metagenomics might provide a better answer. Second, I have a semantic issue: 16S data does not contain ARGs, hence can be used only for prediction of ARGs, while detection of ‘ARG presence’ requires whole metagenomic data. I might be wrong here, and people consider prediction or extrapolation as a presence. Maybe this approach provides a good approximation, but it seems there is no validation or confirmation of this in the paper.
Metagenome cannot be assembled from 16S.
Kind regards,
Igor