Hello Jordan,
I looked into this and I think I have a solution. The summary is that DESeq2 uses the same abundance matrix of raw counts that most tools for community analysis do. Each dataset in a taxmap object is just a table with a taxon_id column and can be accessed with `obj$data$my_table`. You can use those tables the same way the example data set is used in the DESeq2 tutorial here:
To plot information using metacoder, you need per-taxon information. There two ways I can think to get that from DESeq2: 1) run DESeq2 on per-taxon counts. 2) run DESeq2 on OTU/ASV counts and summarize the results per taxon (e.g. mean log fold change per taxon). I will try to demonstrate both in the example attached. I did this rather quickly, so keep an eye out for any mistakes I might have made. I will try to make a more official tutorial in the future when I get a chance.
You will need to download the HTML file and open it with a web browser to see it.
Let me know if you have questions.
Best,
Zach