A structure-function neuronal network model of the rat nervous system

A consensus description of global vertebrate nervous system organization-its basic plan or wiring diagram-has not emerged after centuries of research. Our approach to the problem is based on published experimental neuroanatomical connection data that describe the adult female and male rat nervous system's intrinsic neuronal network organization at the interregional level of scale that is inherited and species specific. This bilateral network has 924 region nodes with a projected 81,582 directed and weighted axonal connections (edges) between them, giving a network density of 9.6%. On average there are 88 output (and input) connections per region, with 31% contralateral connections and 41% reciprocally connected node-pairs. Local network segregation analyzed with multiresolution consensus cluster analysis (based on connection weight) suggests four interconnected first-order systems: a bilaterally symmetric pair associated with behavior control (centered rostrally in forebrain-midbrain), and two bilateral systems associated with behavior execution (centered caudally in rhombicbrain-spinal cord). These four systems then divide into a nested hierarchy of 242 interconnected subnetworks or modules that at lower levels become amenable to experimental manipulation based on strong hypotheses generated from a structure-function model. Global network integration transcending modular boundaries is facilitated by a large rich club, with the "richest club" having 16 highly interconnected hub-pairs, and small-world attributes indicating relatively dense clustering and short path lengths. Computational modeling reveals that focal alterations in node connection weights tend to spread through the network in module-defined subnetworks that are also shaped by node centrality and connection pattern.