Figure 6.

Changes in network activity by pruning MLI → MLI or PKJ → MLI synapses. Simulations were performed where a random set of synapses of either MLI → MLI or PKJ → MLI connections were removed from the model network to investigate the effect of these connection types on the activity of the network. Left column: measurements of MLI population (top row, in red) and PKJ population (bottom row, in blue) firing rates (darker shade) and ISI CVs (lighter shade) when MLI → MLI synapses are randomly pruned by 25, 50, and 75% as well as fully intact (0%) and fully pruned (100%, i.e., no MLI → MLI left synapses at all). Each neuron's mean firing rate and ISI CV was measured over a 60 s simulation of the operation of the network. Solid circles denote the median of the population for each of these statistics. Bars show the first and third quartile to depict the distribution of values across the population. The cross mark denotes the population mean. Right column: similar measurements in the case of PKJ → MLI synapses pruned. Top-left panel: As more MLI → MLI synapses are pruned, the median firing rate of the population of MLIs (dark red) increases due to decreased mutual inhibition. When the synapses are completely pruned, there is very little variance in the population response and quartile bars overlap with the filled circle and are not visible. Additionally, the median ISI CV decreases as more MLI → MLI synapses are pruned (light pink). Bottom-left: As more MLI → MLI synapses are pruned, inhibition onto PKJs from MLIs increases, thus decreasing the median PKJ population firing rate (dark blue) and increases the median PKJ ISI CV (light blue). Top-right: pruning PKJ → MLI synapses has only a subtle effect on the MLI population median firing right and ISI CV. Bottom right: similarly, pruning PKJ → MLI synapses has only a subtle effect on the PKJ population median firing right and ISI CV.