Do we still need to normalize feature table for beta-diversity?

Good afternoon,

I am using the Qiime 2 to analyze the beta-diversity, so just wondering if we still need to normalize the feature table before we put the data in PERMANOVA as we did in Qiime 1?

Thanks very much for the help!

Good afternoon,

Yes! And rarefaction is also possible in Qiime 2.

See this conversation here:


Thank you for your response! I still got a little confused, so we can use the rarefied_table in core-metrics foe both alpha and beta diversity?


Correct. Or you can use the new plugins that come with Qiime 2. :qiime2: Like these!

These plugins perform the rarefaction step multiple times, then combine the result into a unified measurement of diversity.

For PERMANOVA, you could use the rarefied table, or pass in ‘ReadsPerSample’ as an input to the model as another method to control for sampling depth.


I see. Thank you so much for the explanation!

Just a quick correction here - the results are not combined in a unified diversity measurement, but rather, the impact of various rarefaction depths on diversity metrics are summarized and displayed in these interactive visualizations.

The rarefied table is returned as one of the outputs from diversity core-metrics and diversity core-metrics-phylogenetic. You can also rarefy a table directly using the feature-table rarefy command.

Hope that helps! :t_rex: :qiime2:


More clear to me, thank you very much!

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