Fig. 2: Relation between \({{{{\rm{TC}}}}}_{{{{\rm{score}}}}}\) and Q.
From: Leveraging multivariate information for community detection in functional brain networks
A Trend of \({{{{\rm{TC}}}}}_{{{{\rm{score}}}}}\) relative to Q. Each point corresponds to a partition obtained optimizing Q on the HCP FC network, and on which we computed \({{{{\rm{TC}}}}}_{{{{\rm{score}}}}}\) downstream. Colors code the spatial resolution of the partitions indicating the number of clusters. B For each group of partitions with an equal number of modules, we computed the correlation between \({{{{\rm{TC}}}}}_{{{{\rm{score}}}}}\) and Q and reported here the correlation coefficients. Points colored in orange identify statistically significant correlations. C, D. Values of variance of the modules size and symmetry between hemispheres at each spatial resolution for the two partitions with highest and lowest \({{{{\rm{TC}}}}}_{{{{\rm{score}}}}}\). E Example of how the positive correlation observed in (A) is reversed if we consider groups of partitions with equal resolution (or equal number of modules). In this case, we report the reversed correlation for groups of partitions with 6, 7, and 8 modules. We also report a projection on the cortex surface of the modular structure of the 7-module partitions with the highest and lowest \({{{{\rm{TC}}}}}_{{{{\rm{score}}}}}\).
