How to choose correct cutoff for Greengene OTUs

HI!
I am here again with a new question and your kindly explanation of the issue will be highly appreciated
I wonder what will be the differences between the results when we use OUTs 97% and 99% cutoff for training feature classifier?
By that I mean, if the depth of discovered OUTs and strains will be reduced by decreasing the cutoff to 97 …95…so on?

Thanks in advance

Hi!
I have generated a new topic regarding the understanding of the correct OTUs cutoff during training feature classifier , But yet there is no reply!

THE TOPIC WAS AS FOLLOWS:
HI!
I am here again with a new question and your kindly explanation of the issue will be highly appreciated
I wonder what will be the differences between the results when we use OUTs 97% and 99% cutoff for training feature classifier?
By that I mean, if the depth of discovered OUTs and strains will be reduced by decreasing the cutoff to 97 …95…so on?

Thanks in advance

Hi @sajjad.sarikhan - due to the holidays we have been on a reduced schedule - please stay tuned for a reply to your questions!

Hi @sajjad.sarikhan,
Using a feature classifier training set with a higher cutoff value will generally increase sensitivity. A 99% cutoff will contain more unique sequences and more sensitive taxonomic information than lower cutoffs. So I would recommend using a 99% cutoff, in general, unless if you do not expect to need such sensitive information.

I hope that helps!

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.