Document fragment network

[reskyner]

Documentation for

a) the fragment network is used to calculate new molecules in the loader
b) if/how the fragment network is queried to look for new molecules (not found in the loader stage in the back-end)
c) how the cypher queries are wrapped in the back-end
d) how to add new queries as endpoints (some minimal info about the schema needed to allow building of queries)