The effective enhancement of information in 3D small-world networks of biological neuronal cells

Abstract

The cardiovascular system, the kidney, or the brain, are examples of complex systems - where the properties of the systems arise because of the layout of cells in those systems. One way to characterize systems is using networks, where elements and interactions of the systems are represented as nodes and links of a graph. Network’s topology can be, in turn, measured by the small-world coefficient. Small world networks feature increased clustering and shorter paths compared to random or periodic networks of the same size. This suggests that systems with small world attributes can also efficiently transport signals, nutrients, or information within their body. While several reports in literature have illustrated that real biological systems are small-world, yet little is known about how information varies as a function of the small-world-ness () of three dimensional graphs. Here, we used a model of the brain to estimate quantitatively the information processed in 3D networks. In the model, nodes of the graph are neuronal units capable to receive, integrate and transmit signals to other neurons of the system in parallel. The information encoded in the signals was then extracted using the techniques of information theory. In simulations where the topology of networks of nodes was varied over large intervals, we found that in the range information scales linearly with the small world coefficient, with a five-fold increase. Results of the paper and review of the existing literature on model organisms suggest that a small-world architecture may offer an evolutionary advantage to biological systems.