M systems are mathematical models of morphogenesis developed to gain insights into its relations to phenomena such as self-assembly, self-controlled growth, homeostasis, self-healing and self-reproduction, in both natural and artificial systems. M systems rely on basic principles of membrane computing and self-assembly, as well as explicit emphasis on geometrical structures (location and shape) in 2D, 3D or higher dimensional Euclidean spaces. They can be used for principled studies of these phenomena, both theoretically and experimentally, at a computational level abstracted from their detailed implementation. In particular, they afford 2D and 3D models to explore biological morphogenetic processes. Theoretical studies have shown that M systems are powerful tools (e.g., computational universal, i.e. can become as complex as any computer program) and their parallelism allows for trading space for time in solving efficiently problems considered infeasible on conventional computers (NP-hard problems). In addition, they can also exhibit properties such as robustness to injuries and degrees of self-healing.

This paper focuses on the experimental side of M systems. To this end, we have developed a high-level morphogenetic simulator, Cytos, to implement and visualize M systems in silico in order to verify theoretical results and facilitate research in M systems. We summarize the software package and make a brief comparison with some other simulators of membrane systems. The core of the article is a description of a range of experiments inspired by aspects of morphogenesis in both prokaryotic and eukaryotic cells. The experiments explore the regulatory role of the septum and of the cytoskeleton in cell fission, the robustness of cell models against injuries, and, finally, the impact of changing nutrient concentration on population growth.

M 系统是形态发生的数学模型，旨在深入了解其与自然和人工系统中的自组装、自控生长、稳态、自愈和自繁殖等现象的关系。M 系统依赖于膜计算和自组装的基本原理，以及对 2D、3D 或更高维欧几里得空间中的几何结构（位置和形状）的明确强调。它们可以在从它们的详细实现中抽象出来的计算级别上，从理论上和实验上对这些现象进行原则性研究。特别是，他们提供了 2D 和 3D 模型来探索生物形态发生过程。理论研究表明 M 系统是强大的工具（例如，计算通用，即 可以变得像任何计算机程序一样复杂），并且它们的并行性允许以时间换取空间，以有效解决传统计算机上被认为不可行的问题（NP 难题）。此外，它们还可以表现出诸如对伤害的鲁棒性和自愈程度等特性。

本文重点介绍M系统的实验方面。为此，我们开发了一种高级形态生成模拟器Cytos，用于在计算机中实现和可视化 M 系统，以验证理论结果并促进 M 系统的研究。我们总结了软件包并与其他一些膜系统模拟器进行了简要比较。这篇文章的核心是对一系列受原核和真核细胞形态发生方面启发的实验的描述。实验探索了隔膜和细胞骨架在细胞分裂中的调节作用、细胞模型对损伤的稳健性，以及最终营养浓度变化对种群增长的影响。