Currently NFsim takes reactant symmetry into account when computing the total rate of a reaction class, but it doesn't correct when the reaction breaks the symmetry. We (John Sekar and I) believe that this could be fixed easily by considering whether or not the rule transformation breaks the symmetry, instead of on a reaction by reaction basis, as suggested by the NFsim support page. For example, consider an autophosphorylation reaction from the FceRI model:
Syk(Y~U).Syk(Y~U) -> Syk(Y~U).Syk(Y~P) kS
This rule breaks the reactant symmetry, and thus every species containing an instance of this pattern should generate two reactions, or, equivalently, there are two reaction events that are possible with identical rate kS. This can be determined easily by seeing that the transformation is not symmetric with respect to the reactant molecules, which could be determined by applying the transformation to the pattern on the LHS and determining that there is no longer a symmetry.
Currently NFsim takes reactant symmetry into account when computing the total rate of a reaction class, but it doesn't correct when the reaction breaks the symmetry. We (John Sekar and I) believe that this could be fixed easily by considering whether or not the rule transformation breaks the symmetry, instead of on a reaction by reaction basis, as suggested by the NFsim support page. For example, consider an autophosphorylation reaction from the FceRI model:
Syk(Y~U).Syk(Y~U) -> Syk(Y~U).Syk(Y~P) kS
This rule breaks the reactant symmetry, and thus every species containing an instance of this pattern should generate two reactions, or, equivalently, there are two reaction events that are possible with identical rate kS. This can be determined easily by seeing that the transformation is not symmetric with respect to the reactant molecules, which could be determined by applying the transformation to the pattern on the LHS and determining that there is no longer a symmetry.