Selecting a different significance level for core-metrics-phylogenetic package

I am trying to analyze 16S sequence data and am comparing control and test samples to determine whether they are statistically significant. I know that the PERMANOVA significance value is (alpha = 0.05) by default. I am wondering if there is anyway to set the significance value to a different level such as alpha = 0.10 or 0.15. I checked the plugins on the Qiime2 documents page but did not find any argument for this.

Hi there @annenavlekar!

The core-metrics-phylogenetic method in q2-diversity doesn't perform any statistical significance tests - but there are several other actions in that plugin that do. The beta-group-significance does perform PERMANOVA on beta diversity metrics.

Regarding the choice of what is or isn't significant - that choice is completely up to you, and there is nothing you need to change when running these commands, you simply observe the p (or q) value provide by the action, and see if it is above or below the significance threshold that you're interested in.

Now, I am not a statistician, but I highly recommend you consult with one while running these analyses and tests - they are going to be able to help you with crafting statistical design questions, which QIIME 2 can then help you answer.


1 Like

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