Network properties of structural clusters in caenorhabditis elegans nervous system

Abstract

As an anatomical structure for natural computation, a nervous system may consist of discrete computational modules. In addition to developing methods for dissecting those clusters, a large number of studies have examined their characteristics. However none of the previous research identified the clusters of individual neuron-level networks due to the extremely high complexity of the biological organization with multiple agents linked. In this study, we use a wholly computational approach to infer the hierarchical structural clusters of the Caenorhabditis elegans nervous system. Assigning it as an example of a complex network, we reach this goal by using recently developed modularity-based community detection algorithm. After defining the anatomical clusters from its connection diagram (connectome), we investigate the properties of these clusters. The major findings of our analyses are as follows. The spatial properties of the clusters reveal the existence of body-spanning clusters in a biological neural network. We find that the derived clusters show a strong relationship with the worm’s behavioral-level circuits. A possible answer for why the actual system does not follow the minimal wiring principle is presented in the clustered network frame-work.