Common neighbor bias in the different regions of the Blue Brain rat SSCX model

The "common neighbor bias" refers to the tendency of neurons with many graph-theoretical common neighbors to be more likely to be connected to each other. This concept can be split into a bias for afferent and efferent common neighbors.

We quantify it (for a given neuron) as the mean number of common neighbors with connected neurons, divided by the mean number of common neighbors with all other neurons. Further distinction can be made by limiting this analysis to pairs of given neuron types.

In this dataset, we report samples of such common neighbor biases in different regions of the rat SSCX model of Blue Brain. We report biases separately for afferent and efferent common neighbors, and for different pre- and postsynaptic neuron types. Results are lists, where each entry in a list is a result for a single neuron of the indicated pre-synaptic neuron type, quantifying how it affects it connectivity to neurons of the indicated post-synaptic type. A value of 1.0 indicates no bias.