we’re planning a new NGS run, but budgetary restraints ‘force’ us to Ion Torrent …
So, to have a view on the errors associated with the technique, we’ll include (at least) 2 mock-communities (MC): 22 strains and I’m thinking of 1 MC with equal concentrations each and another one with a couple of ‘high abundant’ and some ‘rare’ organisms.
We’ll furthermore resequence some samples from a 454 run, of which some in duplicate, and some new samples in duplicate.
Are there other possibilities I could include to test? Other MC configurations?
Any specific things I should consider when composing an MC?
Ugh, is there really no way to do MiSeq? I have yet to see any IT data that I would want to publish. Anyway…
So we view the MC as a way to assess error rates. With that as the goal, you probably want an evenly distributed MC. I’m not sure what an uneven MC would get you except to perhaps look at PCR bias. Some things that people like to think about include getting similar and dissimilar 16S sequences (e.g. use a couple of Streps and include a Verruco).
Unfortunately not on this project. Next one will definitely be MiSeq.
I was already including 2 spp. from the same genus twice (Deinococcus and Arthrobacter), so that’s ok.
Ok, so forget about the non-equal MC. I’ll have an additional sample resequenced then.