- Poster presentation
- Open Access
Conductance interaction identification by means of Boltzmann distribution and mutual information analysis in conductance-based neuron models
© Santana et al; licensee BioMed Central Ltd. 2012
- Published: 16 July 2012
- Mutual Information
- Boltzmann Distribution
- Stomatogastric Ganglion
- Bivariate Probability
- Boltzmann Probability
Recent research has focused on the causal paths that explain how neuron ionic-conductances interact to produce a particular electrophysiological behavior. Pharmacological blockage methods, neuromodulators and genetic knockouts are among the techniques used to study this issue. In this paper we propose a probabilistic approach based on the computation of the Boltzmann distribution and the mutual information of conductance interactions to learn higher-order, not necessarily pair-wise, potential co-regulation mechanisms from a database of the crustacean stomatogastric ganglion pyloric circuit models. The original database was built from experimental data obtained from lobster stomatogastric neurons [2, 3]. The eight currents in the single-compartment model are based on the lobster stomatogastric ganglion neurons currents.
The basic idea of our work is to assign a probability value p(x) to each neuron model depending on whether or not it satisfies a given electrophysiological property. To assign these values we use the Boltzmann probability distribution, commonly used in statistical physics to associate a probability with a system state according to its energy . From the Boltzmann distribution we compute the mutual information to measure the strength of interaction between a pair of conductances at the time of producing a particular electrical activity.
The particular characteristics of silent neurons were captured in the uneven distribution of bivariate marginal probabilities computed from the Boltzmann distribution. Among all conductance pairs, the highest bivariate probabilities are reached for conductance pairs (gKCa,gKd), (gNa,gKCa) and (gCaT,gCaS). With respect to previous analysis of the group of silent neurons, the correlation (gCaT,gCaS) was the only significant linear correlation identified for silent models in . There were another three conductance relationships reported in  that fitted statistical criteria for correlations, but did not appear to have a linear relationship. They were (gH,gleak), (gKCa,gKd) and (gKCa,gCaT). All these correlations were identified as significant correlations of the computed Boltzmann distribution
Our results show that probabilistic modeling based on the Boltzmann distribution can capture potential co-regulations that are not captured by the correlation analysis. The extension to capture higher-order dependencies between conductances is also straightforward. Furthermore, our results indicate that mutual information analysis allows a more detailed visualization of the structure of the conductance landscape for conductance-based neuron models.
This work has been partially supported by the Saiotek and Research Groups 2007-2012 (IT-242-07) programs (Basque Government), TIN2010-14931, TIN2010-20900-C04-04, Consolider Ingenio 2010 - CSD2007-00018 projects and the Cajal Blue Brain project (Spanish Ministry of Science and Innovation) .
- Hudson AE, Prinz AA: Conductance ratios and cellular identity. PLoS Comp Biol. 2010, 6: e1000838-10.1371/journal.pcbi.1000838.View ArticleGoogle Scholar
- Prinz AA, Billimoria CP, Marder E: Alternative to hand-tuning conductance-based models: Construction and analysis of databases of model neurons. J. Neurophys. 2003, 90: 3998-4015. 10.1152/jn.00641.2003.View ArticleGoogle Scholar
- Turrigiano GG, LeMason G, Marder E: Selective regulation of current densities underlies spontaneous changes in the activity of cultured neurons. J Neurosci. 1995, 15: 1129-1131.Google Scholar
- van Kappen N: The Stochastic Processes in Physics and Chemistry. 1992, North HollandGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.