Minimum Description Length: Application to BK- and Cx43-channels

Morten Schak Nielsen (1), Johannes Pauli Hofgaard (1) and Jakob Kisbye Dreyer (2)

1) Department of Biomedical Sciences and The Danish National Research Foundation Centre for Cardiac Arrhythmia, Faculty of Health and Medical
Sciences, University of Copenhagen, Copenhagen, Denmark
2) Department of neuroscience and pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
(jakobdr@sund.ku.dk)

Abstract:
The mechanistic and molecular basis of many important physiological phenomena can be investigated by analyzing the activity of single ion channels.
Such analysis usually rely on user dependent inputs such as a priori defined Markov models, detection thresholds, or even manual detection of single channel events.  While this may work in some systems, many biological applications involve multiple channels with unknown gating properties. Therefore an alternative unbiased view of ion-channel events is sometimes appropriate. However, idealization of single channel currents in recordings with large number of channels with unknown properties is a difficult problem. Here we develop event detection based on Rissanen's Minimum Description Length principle. We first validate our method on simulated data and then apply it to ion channel systems of increasing complexity, the BK- channel and Cx43 gap junctional channels. Our method reproduces conductances and gating kinetics for these channels. However we also identify complex channel formation in Cx43 indicating that opening of these channels proceeds via a series of sub-steps before full opening occur.