On the GO side, CHEBI has a term "carbohydrates and carbohydrate derivatives" (CHEBI:78616), that could be used to create a new GO term or broaden the scope of an existing one, to lump carbohydrates and derivatives together. Lumping actually fits with the biology. The first thing an animal or S. cerevisiae or E. coli cell (and probably others but I'm ignorant) does upon acquiring a carbohydrate molecule is to phosphorylate it - otherwise it just leaks back out of the cell, so essentially everything we sloppy biologists call "carbohydrate metabolism" is "carbohydrate derivative metabolism" according to nomenclature-obsessed chemists. What makes sense for GO and its users - to accept this distinction and group carbohydrates and carbohydrate derivatives, or ignore it and force our hierarchy to follow "biologist" usage, violating "chemist" usage?
Issues were brought up in https://github.com/geneontology/go-annotation/issues/4017
Suggestion by @deustp01