I suggest to add sort option (for p-value) and cutoff (otu pairs higher than 0.05 p-value can be discarded) for otu.association command. I have over 1500 otus and the list is quite long, so it is hard to manually sort it.
This will be in 1.28.0.
Is there an option for specifying which OTU’s to use in the otu.association command. I have about ~4500 OTUs (15 month temporal dataset), but am only interested in otu.association results for OTUs that make up more than 1% of any given sample or for the ones that appear in >90% of the samples.
Thanks very much,
Probably not essential, but would be a nice feature in the future for otu.association to include options to transform the data? for example otu.association(shared=file.subsample.shared, transform=log(1+x)).
Ameet - you should try filter.shared to filter your shared file as desired.
Thanks very much Pat. Was not aware of this command - makes things very convenient.