Hello and thanks for your amazing work.
In qiime 1 we used to download the values coming from make_distance_boxplots.py to draw box plots in PRISM, is it possible to do it with beta-group-significance or any other function for beta diversity as well?
And also, what is the the threshold of p value and q value for having statistically significant distance with permanova (0.01, 0.05…)?
Thanks a lot,
Great questions! To start with, you can view all the qiime 2 scripts (now called plugins) on this page.
This is the closest thing I could find to make_distance_boxplots:
You could also build these boxplots from the distance metrics, directly from the beta diversity plugins:
As for p-values from permanova, that choice is up to you! Remember to report the R^2 value too; if
Age both have p-values of 0.001, then you can look at the R^2 values to see what has the largest effect size (maybe they are both ‘significant’, but
Age has the larger R2 value == larger effect).
We are working to make raw data used in plots downloadable wherever possible in QIIME2 visualizers. We have an open issue to add this feature to beta-group-significance — but alas there is not a way to do this in the current release. We hope to add this feature soon.
Whatever you want it to be! So the same for any other P-value: you set the threshold you plan to use in your study, and stick with it. A p-value is giving you the probability that the result is statistically significant, so you decide the threshold you want to use.
And @colinbrislawn’s advice about noting the R2 value is important — that’s important to report in publication as it tells you the amount of variation related to each factor.
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.