Hooray, you’re a hero Jamie! This was the issue, and now it seems quite clear. I went ahead a ran the data through the remainder of your tutorial steps after filtering the low frequency features out of the input table. The results are quite interesting, and I’m excited to use this pipeline to test some of the questions we are working on here.
This all occurred because I was working from a table of samples filtered from the full data set, leaving behind features only found in some of the unused samples. I am trying to decide if working on the balances of a sub-set of samples is valid, or if I should handle these samples separately all the way from the start of the analysis.
Anyway, thanks again for all your help. I’m sure I will be back once I’ve broken your algorithms again!
Seth