Hi, I am in a trouble to decide the --p-sampling-depth value.

Metric Sample
Number of samples 45
Number of features 2,146
Total frequency 3,318,807

Frequency per sample

Minimum frequency 30,572.0
1st quartile 55,012.0
Median frequency 70,415.0
3rd quartile 86,541.0
Maximum frequency 136,126.0
Mean frequency 73,751.26666666666

Frequency per feature

Minimum frequency 1.0
1st quartile 11.0
Median frequency 54.0
3rd quartile 264.75
Maximum frequency 402,868.0
Mean frequency 1,546.508387698043

How I to decide the --p-sampling-depth ?



You are not the only one. This is a common question because you are balancing keeping more samples and keeping more reads in each sample.

Mehrbod has a great discussion of this trade-off here:

The good news is that your minimum frequency is pretty good! :+1:

Minimum frequency 30,572.0

You should be able to use --p-sampling-depth 30500 and keep all your samples.


Hi, thanks for you suggestions, however I tried using --p-sampling- depth 30500 but without good results. How can I do?

Tell me more!

How were these results bad? What were you hoping for based on your choice of threshold?

I sent the command :
qiime diversity core-metrics-phylogenetic
--i-phylogeny rooted-tree.qza
--i-table table.qza
--p-sampling-depth 30500
--m-metadata-file sample-metadata.tsv
--output-dir core-metrics-results
I have this results: "There were some problems with the command"

1 Like


Check out the first error listed in your screenshot:

(1/18) Invalid value for '--output-dir': 'core-metrics-results' already exists, will not overwrite.

This directory already exists from your current location on your machine. You can either delete that existing directory, change the name of your '--output-dir' directory, or move to a different location on your machine to create another 'core-metrics-results' directory.

Cheers :lizard:


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