Derived BEL Nanopubs

[ncatlett]

I think the Derived BEL Nanopub concept needs a little work. Two important examples of derived BEL knowledge to consider are orthologization (i.e., the rat and mouse versions of the same BEL statement) and the "inferred direct" combination of unique pieces of knowledge to make a direct statement.

Should these types of derivations each form a new nanopub? How should the annotation be handled in the case that the derivation came from multiple BEL nanopubs? Should these be considered "derived assertions" supported by the original Asserted Nanopub(s)?