Estimation of Channel Kinetics from the Random Fluctuations of Macroscopic Currents

MOFFATT LUCIANO

Tilton School, NH, USA

Conferencia; Ion Channels Gordon Conference; 2006

Gordon Research Conferences

I describe a Maximum Likelihood method for the direct estimation of the kinetic rates from macroscopic currents produced by a population of homogeneous ion channels. As the method proposed by Milescu et al (2005), the distribution of states of the population of channels was approximated by a multivariate Gaussian. Contrary to Milescu et al (2005) work, the local time correlation of the data is taken into account by formulating a continuous implementation of the classical forward-backward recursive algorithm. This algorithm is central to Hidden Markovian Modelling, the standard method for direct estimation of kinetic rates of single channel recordings. A true likelihood function of the macroscopic currents was thus obtained, which allowed to 1) obtain reliable estimates of the standard error of the fitted parameters and 2) compare alternative kinetic models. When tested with simulated data on non-stationary conditions, it was found that reliable estimates of all the kinetic constants, the conductance and the number of channels were obtained. Compared to the previous methods, at the price of a higher computational load, it gives better estimates, reducing the number of traces necessary to obtain a given accuracy by a factor of 4. The method can even give accurate estimates of the kinetic rates in macroscopic stationary conditions, where there is no published alternative method for estimating the kinetic parameters