MCMC Based Selection Probability Estimation for Probabilistic Boolean Networks

An estimation algorithm for selection probabilities of probabilistic Boolean networks is developed in this paper. A recursive estimation procedure is deduced from Bayes theorem and gives the expression of posterior probability distribution of the selection probabilities. To realize the distributions in the fashion of numerical computation, Markov chain Monte Carlo (MCMC) method with Metropolis-Hastings sampler is exploited and provides approximated probability densities. Some numerical examples are illustrated to demonstrate the effectiveness and computation complexity of the proposed framework.