A network of reactions is a commonly used paradigm for representing knowledge about a biological process. How does one understand such generic networks and answer queries using them? In this paper, we present a novel approach based on translation of generic reaction networks to Boolean weighted MaxSAT. The Boolean weighted MaxSAT instance is generated by encoding the equilibrium configurations of a reaction network by weighted boolean clauses. The important feature of this translation is that it uses reactions, rather than the species, as the boolean variables. Existing weighted MaxSAT solvers are used to solve the generated instances and find equilibrium configurations. This method of analyzing reaction networks is generic, flexible and scales to large models of reaction networks. We present a few case studies to validate our claims.
Keywords: weighted MaxSAT, reaction network, reaction constraints