Random forest analysis: Important features as mean decrease in gini

[Jacob]

Hi,
I’m trying to use it to identify the important taxa that differentiate the test and control groups. I noted that the important features and the results are plotted in the form of the mean decrease in accuracy. Is it possible to output the features in their mean decrease in the Gini instead?

In addition for the beta diversity, could the Aitchison(Euclidean) Distance Metric be added as an option too?

I can provide the count tables and meta files if required.

Thank you!

[xia.lab]

Great suggestions! We’ve added both features, and you’ll see them as soon as the servers restart (weekly on Saturday nights)