Identification of hidden Markov models for ion channel currents. II. State-dependent excess noise

For pt.I see ibid., vol.46, no.7, p.1901 (1998). Hidden Markov modeling (HMM) techniques have been applied in the past few years to characterize single ion channel current events at low signal-to-noise ratios (SNRs). In this paper, an adaptation of the forward-backward procedure and Baum-Welch algorithm is presented to model ion channel kinetics under conditions of correlated and state-dependent excess noise like that observed in patch-clamp recordings. An autoregressive with additive nonstationary (ARANS) noise model is introduced to model the experimentally observed noise, and an algorithm called the Baum-Welch weighted least squares (BW-WLS) procedure is presented to re-estimate the noise model parameters along with the parameters of the underlying HMM. The performance of the algorithm is demonstrated with simulated data