ITS SOP and ITS databases

Theory behind mothur
1

Hello

I’ve used the MiSeq SOP to examine 16S data in a couple of projects that we’ve published. I’m now looking at the fungal microbiome in biological samples. This, like 16S, is coming from MiSeq.

I’m even more of a newbie with ITS than I am with 16S, so I’m looking for an SOP, which databases would be best (and how to get to them and make them work with Mothur), etc. Any ideas?

Thanks in advance,

2

Hi Steve
I haven’t written this up as an SOP yet, but here’s my ITS2 batch file. I wrote this before opticlust came out so it’s using average. Also, I don’t chop the reads to the same length-so I should have ensured that the gap extension penalty was 0 but can’t remember if that’s what average in mothur does or not (Pat or Sarah??). Picking an OTU cutoff is pretty questionable for the rapidly evolving ITS, so I’d do both 3 and 5% and see if you are seeing the same patterns in your communities.

3

Hi and thank you! I’ve downloaded your script.

One question. You have this –

#RDP classifier using silva as the reference (similar results as RDP reference but some are classified to species)
classify.seqs(fasta=current, count=current, reference=UNITEv6_sh_dynamic.fasta, taxonomy=UNITEv6_sh_dynamic.tax, cutoff=60)

Once again displaying my newbie status with regard to ITS, but where did you get the UNITEv6 reference database, and how did you fix it so that MOTHUR could use it?

4

And of course when I go to the web it becomes easy (-ish).

I take you used the Mothur download from here: https://unite.ut.ee/repository.php

I see sh.97, sh.99, and sh.dynamic; your code suggests that you used the last one, correct?

I’ll see if I can muddle through from here. My thanks again.

5

oops sorry didn’t change the comment from my v4 code. I use the sh.dynamic because it should be a bit more curated than the hard cutoff ones (ITS doesn’t evolve uniformly across the whole domain).