I have been trying to run the beta-group-significance command using the option "--m-metadata-column Description" and have been getting this error message:
"All values in the grouping vector are unique. This method cannot operate on a grouping vector with only unique values (e.g., there are no 'within' distances because each group of objects contains only a single object)."
I understand that there is a bug with the "pairwise" parameter - I have read other posts on this forum that have been resolved by filtering down the distance matrix and running the beta diversity command without the pairwise parameter. I have tried doing this (filtering the distance matrix to just the 2 samples I am trying to compare) and still the same error message.
Is this due to the bug, or am I misunderstanding the beta diversity command? Is it not possible to use this command to compare two unique samples?
Here are links to similar issues that have been resolved. Any input would be greatly appreciated.
If your description is unique, you can’t run permanova on that metadata column. Permanova requires multiple samples per group, so it can compute within group variance (within group distance) and between group variance (between group distance). It’s behaving exactly the way it should when you try this.
Id suggest either breaking your discription down into pieces, for for instance, if your description is cage x treatment for a mouse experiment (i.e. m1c2t0), then I’d break it into a column that is an subject identifier (mouse), a cage identifier (cage) and a treatment identifier (treat), so mu new file looked liek this: