Rare and abundant OTU

Commands in mothur
1

Dear Mothur Group,

I have similar question like many others have in the forum.
I want to remove the singletons or say cutoff of 10 before running classify.otu and make.shared command. In order to do so I did following step.

split.abund(fasta=Final.subsample.fasta, list=Final.subsample.an.list,group=Final.subsample.groups,cutoff=10)

From above command I got lots of output files named rare and abundant.
So my question is, if its okay to use either rare or abundant file as an input to make shared file and classify my OTU’s only based on rare or abundant.
But in biological world rare otu if it belongs pathogenic bacteria plays a crucial role and thus i am bit skeptical here.
At the same time singletons may give a very wrong picture of Diversity plots. So it would really kind if someone shed light on this topic to me.

Thanks and looking forward for replies!!

2

I’m anti removing rare OTUs so I would tell you that anything you do along these lines is a bad idea :slight_smile:

3

Thank you for your reply~~
That means you dont suggest to remove singletons…
but aren’t singletons a problem for diveristy plots? for example out of ten biological replicates its expressing just in one sample and that too one time how can one justify that.

Thanks
Best,
Rich

4

That means you dont suggest to remove singletons…

Correct.

but aren’t singletons a problem for diveristy plots?

Not sure how that would be. If you have no singletons then you’re essentially saying that you have sequenced everything twice, which is unlikely. Also, if you remove singletons from a library with 4000 reads, what’s your threshold for a library with 10000 reads?