Include/Exclude NA values in Categorical Clinical Data to get idea of data availability in each group

[inodb]

Maybe we can use a checkbox like we do for the study view to include/exclude NA like values (e.g. "Unknown')

Note that in the statistical testing we don't have a good way to share NA like values between frontend/backend. Both are hardcoded. It might be good to specifically pass the NA like values to the enrichment api endpoint from the frontend to avoid this. Or we could add it as a property in portal.properties or something. Could lead to strange results if the testing uses a different set of NA values than the frontend (EDIT: we decided to add it as a property to portal.properties and manage it at the instance level)