Striatal network modeling in Huntington’s Disease
Fig 4
2500 cell network model shows transition as parameters are varied.
Several quantities calculated from 200 second network simulations divided into 5, 40 second long, non-overlapping segments. Quantities shown (except c, inset) are calculated from single unit spike trains averaged across all active cells in each 40 second segment. Bars show SEM in these averages across the 5 segments. Results are versus inhibition, gI, at several levels of excitation, gE (see keys). Simulations at different levels of inhibition, gI, are different realizations of random networks with given connection probability of 0.2. Simulations at any given level of inhibition are identical except for the different levels of excitation, gE. (a) mean firing rate, r, (a, inset) rescaled mean firing rate, r(gI/gE), (b) mean ISI CV, (c) maximum likelihood lognormal distribution shape parameter, σLN, (c inset) number of active cells (which fire at least one spike in the whole 200 second network simulation period), (d) KS distance between the model generated ISI distribution and the estimated maximum likelihood exponential distribution (KSE, brown), γ distribution (KSγ, orange), lognormal distribution (KSLN, pink) and inverse gaussian distribution (KSIG, cyan) for two different values of gE (gE = 40, solid and gE = 50, dashed). (a-c) Black, red, green and blue circles indicate the simulations described in Fig 5. (a-d) Black (WT) and red (HD) arrows (and boxes in (d)) indicate the two individual network simulations which are best-fit for two particular restricted age experimental datasets, denoted ‘WT75’ and ‘HD12’. The ISI statistics for these two simulations and the two corresponding experimental datasets are shown in Figs 8 and 9.
doi: https://doi.org/10.1371/journal.pcbi.1007648.g004