It will often happen that I want to run the exact same analysis or just some parts of it but with my samples organised in different groups this time.
It would be nice if mothur had command like regroup(group=jadajada.groups, A[11A-12A-13A]-B[11B-12B-13B]) for creating new groupfile so that to rerun some parts of your analysis with differently grouped samples you dont have to go back to the beginning to create your group file or write a script to create a group file.
To think of it the functionality might even be build in to some commands - like venn(groups=A[11A-12A-13A]-B[11B-12B-13B])
In illustration lets say we have patients(nr.-s), treatments(letters) - we might want to run the test with grouping samples (11A;15C etc.) by patient or by treatment.
Thanks for the suggestion. We’ll put it in the queue.
I came across this post from three years ago in the forum when I was searching for “mothur regroup”. I did a study with litterbags and, in my oligos file, I have things barcoded by each litter bag. However, there are some analyses I’d like to perform at the experimental plot level and I haven’t figured out a way to go back and tell mothur, “these 5 litter bags all belong to plot 1, these 5 to plot 2, etc.” Is there something akin to a regroup function presently available? If so, what is the command called?
I think what you’re looking for is the design file, which we use in a variety of beta-diversity tools such as amova, homova, metastats. If you want to group samples for doing alpha diversity (probably a bad idea) you could always use the merge.groups command.