Is it necessary to set p-perc-identity the same to the level of reference sequences in vsearch cluster-features-closed-reference

[HeinzG (Heinz)]

Hello,

I plan to use qiime vsearch cluster-features-closed-reference to pick up OTUs using the green_99otus.qza as reference sequences, but I am wondering if it is necessary for me to set --p-perc-identity to 0.99. Could I just set --p-perc-identity to 0.97 or 0.90 when I use green99_otus.qza as reference sequences. Just like below:

qiime vsearch cluster-features-closed-reference
–i-table table.qza
–i-sequences rep-seqs.qza
–i-reference-sequences green_99_otus.qza
–p-perc-identity 0.97 \ #or 0.90
–p-strand “both”
–o-clustered-table table-cr-green-99.qza
–o-clustered-sequences rep-seqs-cr-green-99.qza
–o-unmatched-sequences unmatched-cr-green-99.qza \

Thanks for your time!

Heinz

[jwdebelius (Justine Debelius)]

Hi @HeinzG,

It is not necessary! Large databases do clustering to reduce the database size - thus the 97% or 99% identitiy. You can cluster against they at any identity you choose. However, I think it’s recommended to match the two.

Best,
Justine

[HeinzG (Heinz)]

Thank you for your help!