OTU tables from data that spans multiple years

[-Rf]

Samples collected for a time series that spans multiple years results in several open reference OTU tables.  To combine these I converted the open ref to closed ref and then used the merge_otu_tables.py script.  I am just wondering if anybody else has a suggestion for a more elegant way to handle time series data rather than just converting to reference based, losing all of the New.Reference & New.CleanUp.Reference OTUs and then combining? Thanks!

[Colin Brislawn]

Good morning,

The only way to combine open-ref tables is to throw away all the de-novo OTUs (so that these are closed-ref tables), then combine the closed-ref tables of OTUs. Because of the way de novo OTUs are created, combining them is not feasible. 

When I've worked with data from multiple runs over a series of months, I usually pool all my data (old and new) then pick OTUs on this combined set. While reprocessing data is a little annoying, it let's me preserve all novel OTUs in my samples.

I hope that helps! 

Colin