We’re getting to the point where the sequencing of replicates is looking more and more feasible. I’d like to see an analysis of similarity (ANOSIM), MANOVA, or AMOVA-type module built in to Mothur. It should be a fairly simple step (she says naively) beyond building list file. The way that I’m envisioning the work flow would be:
- (…and the steps leading up to it) delineating OTUs over a data set that comprises replicated sequence libraries over multiple experimental treatments/conditions
- As we do already, a a groups file would need to be provided, but in addition, we would also need to provide a file that defines how the libraries fit into the replication “scheme”
- The list file would need to be formatted as a matrix of samples vs OTU counts (or relative abundances)
- Calculation of the F statistic (at all distance levels or at a user-specified distance level)