Get working BioCyc model

[djinnome]

To minimize the curation effort required to get a working model in RhodoCyc, we start with the current FBA model (Joonhoon's model), and we add a penalty to the objective for every reaction that does not have a biocyc mapping. We then run FBA and the number of reactions without biocyc mappings that have nonzero fluxes should be close to the minimum number of reactions that need to be mapped to get a working FBA model.

@ShantMahserejian does this make sense?