Set up a process and API to identify carbohydrates

Presently we rely on manually curated list created by Matt. This needs to be improved!

Ideally a process which is a part of PDBeChem pipeline is implemented to find sugars and stores results in the database.