The effects of different parameters on the coupling coefficients measured between neurons in a passive population model.

<p>In (A-C) a single parameter is varied (two values are shown) and the coupling between model neurons at different distances was measured for 100 generated networks for comparison with the coupling between pairs of dINs measured experimentally (shown by dots in each graph; [<a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1004240#pcbi.1004240.ref029" target="_blank">29</a>]). The notched grey bars show the median and 25% to 75% quartiles of the model coupling coefficient at each distance. The lines represent the 5% to 95% percentiles. Starting parameters (see text): R_i = 80 Ωcm, R_GJ = 600 MΩ, min-dist = 20 μm, max-dist = 70 μm, g_lk = 0.25 mS cm^-2 and axon diam = 0.4 μm. In each case the visual match to experiments is better in the left column than in the right column. (A) R_GJ reduced from 600 to 300 MΩ or increased to 900 MΩ. (B) Axon diameter increased from 0.4 to 0.6 μm or reduced to 0.2 μm. (C) Minimum distance from the caudal soma was reduced from 30 to 10 μm or increased to 40 μm. The parameters min-dist and max-dist correspond to the left and right-hand edges of the GJ distribution step in <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1004240#pcbi.1004240.g002" target="_blank">Fig 2D</a> respectively. When gap junctions were distributed closer to the somata of the neurons, the coupling was stronger (in part because there are more gap-junctions in the network in total).