It's not really possible to do in schema- however you can add both joins and dynamically select them at point of use, e.g. with js initialization of query activity.
I once worked on a project that had a table composed of hundreds of unlabeled cols that were ~grouped in sets, e.g. column1+column318+column319+column618 would be one of many sets defined in an Excel doc.
Made it sort of manageable by using js init to select which groups to query and their relationships to other tables.