1: \begin{abstract}

2: 

3: In this paper we address the problem of Monte Carlo approximation of

4: posterior probability distributions in stochastic kinetic models

5: (SKMs). SKMs are multivariate Markov jump processes that model the

6: interactions among species in biochemical systems according to a set

7: of uncertain parameters. Markov chain Monte Carlo (MCMC) methods

8: have been typically preferred for this Bayesian inference problem.

9: Specifically, the particle MCMC (pMCMC) method has been recently

10: shown to be an effective, while computationally demanding, method

11: applicable to this problem. Within the pMCMC framework, importance

12: sampling (IS) has been used only as the basis of the sequential

13: Monte Carlo (SMC) approximation of the acceptance ratio in the

14: Metropolis-Hastings kernel. However, the recently proposed nonlinear

15: population Monte Carlo (NPMC) algorithm, based on an iterative IS

16: scheme, has also been shown to be effective as a Bayesian inference

17: tool for low dimensional (predator-prey) SKMs. In this paper, we

18: provide an extensive performance comparison of pMCMC versus NPMC,

19: when applied to the challenging prokaryotic autoregulatory network.

20: We show how the NPMC method can greatly outperform the pMCMC

21: algorithm in this scenario, with an overall moderate computational

22: effort. We complement the numerical comparison of the two techniques

23: with an asymptotic convergence analysis of the nonlinear IS scheme

24: at the core of the proposed method when the importance weights can

25: only be computed approximately.

26: 

27: %Insert your abstract here. Include keywords, PACS and mathematical

28: %subject classification numbers as needed.

29: \keywords{Nonlinear population Monte Carlo \and particle Markov

30: chain Monte Carlo \and sequential Monte Carlo \and stochastic

31: kinetic models}

32: %% \PACS{PACS code1 \and PACS code2 \and more}

33: %% \subclass{MSC code1 \and MSC code2 \and more}

34: 

35: \end{abstract}

36: