Investigating the properties of bio-chemical networks of artificial organisms with opposing behaviours

摘要

Organisms, be it singled-celled organisms or multi-cellular organisms, are constantly faced with opposing objectives requiring different sets of behaviours. These behaviours can be classified into two, predatory behaviours or anti-prey behaviours, with one set of behaviours causing an opposite effect to the other. A healthy organism aims to achieve its equilibrium state or to be in homeostasis. Homeostasis is achieved when a balance between the two opposing behaviours is created and maintained. This raises some questions: is there an innate mechanism that encodes for these categories of behaviours? Is there also an innate mechanism(s) that resolves conflicts and allows switching between these two opposing behaviours? If we consider artificial organisms as single-celled organisms, how do the organisms' gene regulatory network, metabolic network and/or signalling network (their biochemical networks) maintain homeostasis of the organisms? This paper investigates the properties of the networks of best evolved artificial organisms, in order to help answer these questions, and guide the evolutionary development of controllers for artificial systems.