This paper presents a real-time simulation involving ''protozoan-like'' cells that evolve by natural selection in a physical 2D ecosystem. Selection pressure is exerted via the requirements to collect mass and energy from the surroundings in order to reproduce by cell-division. Cells do not have fixed morphologies from birth; they can use their resources in construction projects that produce functional nodes on their surfaces such as photoreceptors for light sensitivity or flagella for motility. Importantly, these nodes act as modular components that connect to the cell's control system via IO channels, meaning that the evolutionary process can replace one function with another while utilising pre-developed control pathways on the other side of the channel. A notable type of node function is the adhesion receptors that allow cells to bind together into multicellular structures in which individuals can share resource and signal to one another. The control system itself is modelled as an artificial neural network that doubles as a gene regulatory network, thereby permitting the co-evolution of form and function in a single data structure and allowing cell specialisation within multicellular groups.

翻译：本文提出了一种实时仿真框架，系统内"类原生动物"细胞在二维物理生态系统中通过自然选择进行演化。选择压力源于细胞需从环境中获取物质与能量以完成分裂繁殖的约束。细胞并非天生具有固定形态，它们可利用自身资源构建功能性节点附着于表面，例如感光受体（photoreceptors）或运动鞭毛（flagella）。关键之处在于，这些节点作为模块化组件通过输入输出通道与细胞控制系统相连，使得进化过程能够在利用通道另一端预发育控制通路的同时，将一种功能替换为另一种。其中值得关注的一类节点功能是黏附受体——它允许细胞结合形成多细胞结构，个体间可共享资源并传递信号。控制系统本身被建模为兼具基因调控网络功能的人工神经网络，从而在单一数据结构中实现形态与功能的协同进化，并支持多细胞群体内的细胞特化。