Journal article
Robust inference and modelling for the single ion channel
Journal of Statistical Computation and Simulation, Vol.75(7), pp.513-529
2005
Abstract
Statistical modelling and inference for a single-ion channel have principally been carried out using finite-state space continuous-time Markov chains. Statistical inferences for the closed and open dwell times and the kinetic rate constants between states have then been arrived at via maximum likelihood methods, including the use of the EM algorithm. The fundamental assumption behind this theory is that one has the correct number of closed and open states, something which may not be easily determined by the use of current methods for modelling, say, the number of components in a mixture of exponential distributions used to fit, say, the 'closed' dwell times. Here, we show that the use of a robust L2 estimator can outperform the EM algorithm both when the correct number of states is apparent and also when there are small deviations from the supposed models. After describing the statistical models used to demonstrate these results and how they lead to particular mixtures of exponential distributions the comparison is then made between the performances of the estimators (robust L2 and maximum likelihood (via the EM algorithm)). The resulting performances in terms of means and standard errors of the estimated kinetic rate constants are then assessed. The estimating equations derived from minimizing the L2 distance are given explicitly in the appendices.
Details
- Title
- Robust inference and modelling for the single ion channel
- Authors/Creators
- B.R. Clarke (Author/Creator) - Murdoch UniversityP. McKinnon (Author/Creator) - Murdoch University
- Publication Details
- Journal of Statistical Computation and Simulation, Vol.75(7), pp.513-529
- Publisher
- Taylor & Francis
- Identifiers
- 991005544992007891
- Copyright
- 2005 Taylor & Francis Ltd
- Murdoch Affiliation
- School of Chemical and Mathematical Science
- Language
- English
- Resource Type
- Journal article
UN Sustainable Development Goals (SDGs)
This output has contributed to the advancement of the following goals:
Source: InCites
Metrics
34 Record Views
InCites Highlights
These are selected metrics from InCites Benchmarking & Analytics tool, related to this output
- Citation topics
- 1 Clinical & Life Sciences
- 1.79 Molecular & Cell Biology - Physiology
- 1.79.239 Ion Channelopathies
- Web Of Science research areas
- Computer Science, Interdisciplinary Applications
- Statistics & Probability
- ESI research areas
- Mathematics