I have a very short question: I am using the q2picrust2 for predicting the functional attributes of my dataset (I prefer picrust2 to Tax4Fun and Piphillin because it is more well-cited): I am using the code on the q2 picrust2 Tutorial · picrust/picrust2 Wiki · GitHub
I read that PICRUST only works for GG database, however, from the code, I realize there is no taxonomy.qza file as input, so does it matter if I have used SILVA or GG for my taxonomic annotations? I mean, the functional attributes for the various samples should be independent of taxonomic attributes, no?
A separate question: Can anyone here please suggest some good visual graphic tools to display functional data? I am afraid I am inexperienced in this, so some help would be really appreciated
One of the reasons PICRUSt is well cited is because there was an earlier veersion of the algorithm which people are familiar with.
PICRUST 1 worked with Greengenes because the phylogenetic inference required a tree and at that time, closed reference OTU clustering was a way to get a consistent mapping to a tree for phylogeny. Since the method was published, phylogenetic placement algorithms have ben developed for marker gene data, meaning you can get phylogenetic inference that way. And, even with PICRUSt 1, a lot of people would run open reference clustering on their data against Silva and then use closed ref for PICURSt.
You could always try things like a bar chart (“taxa” may be a misnomer") if you have a name or groupings for the functions.