Hi all,

I have a question about the Good’s calculator: In order to calculate coverage, the equation uses the number of OTUs that have been sampled once. That suggests that in order to calculate coverage, the sequence data used cannot have had the singletons removed in advance? How about subsampling or rarefaction? I have seen in papers the coverage being calculated before and after the removal of singletons. What would be the recommended data treatment to be used for Good’s calculation?



Correct. Removing singletons is a really bad idea. Use rarefaction for your alpha and beta-diversity analyses and subsampling for your OTU-based analyses.