Chasing the dreams of early connectionists

Abstract

Mapping and examining the wiring pattern of neural systems is a fundamental pillar of neuroscience. In this Viewpoint, we review a recently described mesoscale connectome map of the mouse brain. We underscore the map's high spatial resolution and discuss key organizational network attributes of the presented connectome, its potential impact on neuroscience, and the general importance of connectome maps to obtain insight in the workings of the brain at a system's level.

Figure 1

(A) Circular representation of the 25% strongest intrahemispheric connections of the mouse connectome as mapped by Zeng and colleagues, with nodes arranged according to community participation (6 main communities are observed). Hub-to-hub connections (red) are shown on top. (B) Schematic figure of a sagittal slice of the mouse brain taken from the Allen Brain atlas (www.brain-map.org). (C) The mouse connectome reveals a high level of topological clustering (> random level) and short communication pathways (∼ random level), indicative of a small-world organization. (D) Preliminary findings suggest the presence of densely mutually interconnected hub nodes (red points in A and B, left histogram) with hub-to-hub connections (red edges in A) that span on average longer physical distances (middle), comprise higher ranked weights and are more often part of a bidirectional pathway (right) compared to the other pathways in the network (gray edges in A).

Figure 2

Neural networks can be described and examined as a set of nodes (e.g., neurons or mesoscale brain sites) and a collection of connections between nodes (e.g., reconstructed anatomical pathways) (most left panel). Within this mathematical framework, information on the local organization of the network can be, for example, provided by the level of network clustering, reflecting how strong the neighbors of a node are connected (left panel, orange), and the extent to which nodes form local subnetworks or communities (left panel, blue). Information on the global organization of a network can, among other measures, be examined in terms of the average number of steps that is needed to travel from one place to another place in the network (right panel, green). Due to their high level of connectivity and central embedding in the network, hub nodes and their connections often form a prominent structure within the overall network (most right panel, red).