Bayesian Verification of Chemical Reaction Networks

ABC(SMC)$$^2$$: Simultaneous Inference and Model Checking of Chemical Reaction Networks

“…Fig. 4 in the Appendix) introduced in [36], which addresses the following problem. Consider a data generating stochastic system S (in this work, a CRN), where we denote the generated data as y obs .…”

“…A complementary result can be drawn over Θ ¬φ . The full procedure and further details are presented in [36] and summarised in Appendix B. The limitations of the Bayesian Verification framework of [36] lie in the parameter synthesis part.…”

“…The full procedure and further details are presented in [36] and summarised in Appendix B. The limitations of the Bayesian Verification framework of [36] lie in the parameter synthesis part. Parameter synthesis of pCTMCs is considered in the work of [11], and accelerated by means of GPU processing in [12].…”

“…We address the scalability limitations of our previous work [36], and specifically the parameter synthesis part: in [36] the synthesis was calculated symbolically, which practically limited the applicability to CTMCs with small state spaces and a few parameters. We incorporate here statistical model checking within the Bayesian inference framework and instead estimate parameter regions.…”

“…On the contrary, here we work with partially observed data and stochastic models with intractable likelihoods, and must rely on likelihood-free methods and statistical parameter synthesis procedures. Building on previous work [36], which allowed for likelhood-free Bayesian Verification of systems, the presented framework is applicable to a wider variety of stochastic models.…”