table taxonomy with cut-off of frequency

[rparadiso (Rubina)]

Good morning,
I want you know if it is possible to obtain a table of frequency where I can see only the taxa with a frequency > 0.5% in at least one sample?
If yes can you show me the script?

thank you so much

Rubina

[Mehrbod_Estaki]

Hi @rparadiso,
You’ll want to check out the filtering tutorial here for some examples, including what you describe, the only catch is that at the moment QIIME 2 only does abundance filtering and not relative-abundance filtering. So you’ll have to do some quick calculations to figure out the .5% in abundance.

[rparadiso (Rubina)]

so I have to use the script

qiime feature-table filter-features
–i-table table.qza
–p-min-frequency 10
–o-filtered-table feature-frequency-filtered-table.qza

And my --p-min-frequency is?

Thanks

[Mehrbod_Estaki]

Hi @rparadiso,
You’ll want to manually calculate what .5% of your total reads is and place that number there. You can find your total reads data in the table summary vizualization.
Also, based on your initial request, you’ll want to add this too:

--p-min-samples 1 \ # or whatever minimum # of samples you want it to be

One thing to point out that these filters are applied to the whole dataset, so the minimum .5% means a feature must be found x times across ALL of your samples and not just one. I’m not aware of an easy way to filter based on per/sample as you describe in QIIME 2.