Test whether BNG and Kappa exports produce identical network-free dynamics

[lh64]

There are a few differences between BNGL and Kappa that could lead to different dynamics if not properly accounted for. For example, BNGL includes the dot operator that can distinguish between a bimolecular and unimolecular binding event, e.g., A(a) + A(a) -> A(a!1).A(a!1) vs. A(a).A(a) -> A(a!1).A(a!1). The first rule explicitly requires that the two reactants be within separate complexes, while the second rule requires that they be within the same complex (i.e., a ring closure event). The dot operator is not supported in Kappa, so both rules would be written as A(a) , A(a) -> A(a!1).A(a!1). However, Kappa allows two rate constants to be defined to account for the two different cases accounted for by the dot operator in BNGL. In PySB, the BNGL rules can be written as A(a=None) + A(a=None) >> A(a=1) % A(a=1) and A(a=None) % A(a=None) >> A(a=1) % A(a=1). It's not clear to me how these two different rules are exported to Kappa and whether NFsim and Kappa simulations would match in this case.