Correction: Bayesian parameter estimation for dynamical models in systems biology ( PLoS Comput Biol 18:10 (e1010651) DOI: 10.1371/journal.pcbi.1010651)

Abstract

In section 1.6 Constrained interval unscented Kalman filter Markov chain Monte Carlo (CIUKF-MCMC) of the Materials and Methods section, there is an error in Theorem 1. Specifically, in several of the equations in the theorem, some of the indices on x and y in the exponents are incorrect. Please find the correct theorem below; Theorem 1 (Marginal likelihood (Theorem 1 of [26] and 12.1 of [67])) Let yk denote the set of all observations up to time tk as defined in Section 1.2. Let the initial condition be uncertain with distribution p (x0|θ). Then the marginal likelihood is defined recursively in three stages: for k = 1,2,. . . 1. Predict the new state from previous data (Formula Presented) 2. update the prediction with the current data ( Formula Presented )3. and marginalize out uncertainty in the states( Formula Presented )