I’m quite new to mothur, and I’m probably missing something obvious. So far I have followed SOP, and things went well. Mothur is great software which makes my understand how stuff works.
I got to classify.seqs step.
The example abrecovery.taxonomy in classify.seqs looks like that:
The taxonomy files I have got from taxonomy outline page contain classification to the genus level.
As far as I understand custom taxonomy files can be relatively easily made, but can I find taxonomy files with more levels somewhere?
Can you help?
The problem here is trying to determine bacterial species based on a short (~300bp) sement of 16S rDNA. We’ve tried it many times with different methods, it just can’t be done reliably. We estimate that when 454 gives us 600bp-reads, species-level classification will be a possibility.
First off - the sequence you have here went to the species level… As peleliu mentioned length is a big factor. But you also have to have species names in your reference taxonomy file and you need to have multiple sequences per species to make it work. The RDP training set only goes to the genus level and the greengenes training set only goes to the species level for a handful of species. Remember that what you (and everyone else) is trying to do is infer something about 5,000,000 bases from <500. :?
Thank you very much for fast response.
Is till have a question what is your cooment, if different OTUs of the same genus significantly differ among groups in your experiments?
Do we call them Xbacterium OTU1 and Xbacterium OTU2, and it’s too unreliable to find out what OTU1 and OTU2 may represent?
We refer to them as OTU 1, 2, 3, etc and indicate the consensus taxonomy for that OTU. You will certainly have many taxa with multiple OTUs.