Integrate all against all cognate ligand comparisons to procoggraph

[m-crown]

We use PARITY similarity to compare ligands within an EC class for a bound entity. Finding the most similar cognate ligands to a cognate ligand would also be useful - however could prove to be a very time consuming one-time exercise (could test running on a subset and get mean run time to extrapolate).

This could prove useful for exploring potential for promiscuity in enzymes.

[MattBashton]

Also possibility of user search with a mol file of interest.