I have been using my subsampled shared file to load into other statistical software (like Primer) and then I read on here how the subsampling command only runs one round of sampling whereas rarefying will subsample sequences multiple times (i think i saw the default as 1000x) and then averages the counts.
I am wondering, is there a way I can subsample 1000x and generate a shared file that I can load into Excel or another program?
You would have to generate 1000 shared files. What are you doing in primer that you can’t do in mothur? summary.single will rarefy the data for you for alpha diversity (subsample=T) and dist.shared (subsample=T) will get you the rarefied beta diversity distance file.
I’ve thought about how to do this as well because it’d be nice to have a repeatedly subsampled shared file for things like indicator species analysis or species distribution modeling. But I and the few statisticians I’ve talked to haven’t come up with a way to subsample without also hitting the 0 doesn’t mean 0 issue. The samples with more sequences will always have more OTUs because we are sampling them more intensively. I can’t think of a way to repeatedly subsample that doesn’t inflate detected richness which is the problem that we’re trying to fix with subsampling.