Fig. 2: Assortative mixing.
From: Assortative mixing in micro-architecturally annotated brain connectomes
a Given an annotated network where each node has a local annotation score, we can quantify the tendency for nodes with similar scores to be connected using the assortativity coefficient. This coefficient is defined as the Pearson correlation between the scores of connected nodes42. This relationship between the scores of connected nodes can be visualized with a scatterplot of a network’s edges where the position of each edge is determined by the annotation scores of its two endpoints. Here, the intersection of the two dashed lines indicates the position of the edge highlighted in the zoomed-in frame of the network. In this example, the assortativity coefficient (r) is equal to 0.54. b To control for spatial constraints, the assortativity coefficient of an empirical annotation is compared to the assortativity coefficients of n = 10,000 null annotations that preserve the spatial autocorrelation of the empirical one38,43,44. The boxplots in b represent the 1st, 2nd (median) and 3rd quartiles of the null distribution; whiskers represent endpoints of the distribution. The spatial coordinates used for the visualization of the connectome corresponds to the parcel centroids of the 800-nodes Schaefer functional atlas89.
