Title: Adaptive design of multi-functional network in a primitive organism

Toshiyuki Nakagaki

Short Abstract
Transport networks are a ubiquitous feature of both social and biological systems. In each context, there is a complex trade-off between the cost of making and maintaining the network, its transport efficiency and its fault tolerance. Traditionally, the emphasis in man-made infrastructure networks has been to balance high efficiency with low cost. More recently, the consequences of accidental network failure or increased risk of deliberate attack, has focused attention on development of more robust network architectures. However, finding the most appropriate solution to such a combinatorial optimization problem is not straightforward. Here we show that the true slime mold {¥it Physarum polycephalum} can be used as a model experimental system to gain insight into the rules governing de-centralized, self-organized, adaptive network development. This simple biological system can establish networks with comparable efficiency, fault-tolerance and cost to real-world infrastructure networks, in this case judged in comparison to the Tokyo rail system. We argue that the solutions reached by such biological systems have been honed by many cycles of evolutionary selection pressure and are likely to yield a reasonable balance between cost, efficiency and resilience. Furthermore such systems develop without centralized control and may represent a readily scalable solution for growing networks in general. We have therefore developed a biologically-inspired mathematical model that captures adaptive network formation in {¥it Physarum} with a minimal set of equations. We anticipate that this model encapsulates the core mechanisms needed for adaptive network development, and should have wide applicability to guide network construction in other domains.