Hi all,
I have a data set consisting of mutations that are either deleterious or nonharmful, and from each instance I have calculate a p-value through some empirical distribution. I was wondering if it is possible for me to plot my p-values in an AUC or do I need some training and test data to be able to accomplish that or is it sufficient enough to set up a significance level?
I know that the area measures discrimination, i.e., the ability of the test to correctly classify those with and without the disease-assocation. But I'm not sure if I can or how I could apply AUC to my problem? Where do I start?
Best,
Sofia