Determining the rules for self-organised adaptive biological networks

Mark Fricker (University of Oxford)

Physical transport networks occur at many different scales in biological systems, ranging from the endoplasmic reticulum at a sub-cellular level, to vascular networks operating within plants and animals, to complete organisms spanning hectares in the case of woodland fungi. We have been developing tools to extract and characterise the biological network formed in each of these contexts, using both domain-specific and general graph-theoretic metrics, although the latter have not yet yielded much in the way of useful insights for these essentially planar spatial networks. Here we focus on fungal networks, as unlike vascular systems, they are not constrained to a predictable structure, but explore space in the search for patchy and ephemeral nutrients, continuously adapting to varying external conditions, in the face of competition, damage and predation. Exploration, repair and combat require internal transport of nutrients from spatially disparate sources to these rapidly altering sinks. Thus, the network architecture and internal flows continuously adapt to local nutritional cues, damage or predation, through growth, branching, fusion or regression. As these organisms do not have any centralized control system, we infer their relatively sophisticated behaviour emerges from parallel implementation of many local decisions that collectively manage to solve this dynamic combinatorial optimization problem. To understand how such behaviour is achieved and coordinated, we have developed combined imaging and modelling approaches to characterize the network structure, link the structure to predicted nutrient transport, based on models of fluid flow dynamics, and then test these predictions using measurement of nutrient flows using photon-counting scintillation imaging. In parallel, we have explored control of network development in the acellular slime mold, Physarum polycephalum, which is taxonomically unrelated to the fungi, yet appears to exemplify common solutions to self-organised adaptive network formation driven by fluid flows, local rules and oscillatory behaviour. The simplicity of the bioinspired Physarum model hints at a class of algorithms that give quasi-optimal solutions to balancing cost and transport efficiency using a combination of iterative local rules with long-range coupling. By adopting a comparative approach across widely divergent organisms, we aim to identify universal biological algorithms that yield optimized network design.