Complexity Science: Systems Thinking from New Biology to New ICT Challenges

我们建议在复杂性科学中开设六门为期一周的授课课程。“复杂性”在这里是在一个特殊的意义上使用的，指的是具有大量相对简单的部分的系统，这些部分相互作用产生集体行为（这可能不是从单个部分的行为中显而易见的）。虽然历史上复杂性科学研究的系统就像沙堆中颗粒的相互作用一样统一，但这类系统最丰富的例子是生物系统：例如由细胞组成的有机体，或由相互作用的有机分子网络组成的代谢系统。人造系统中有趣的例子包括万维网内容或城市发展模式的某种自我组织特性。这些课程的主题将是计算/信息技术和生物科学的复杂性原则的统一。这将解决这些学科面临的新挑战：1）现代信息和通信技术（ICT）系统需要灵活，强大，适应性强和可扩展，但传统的设计，控制和软件工程方法越来越无法提供这一点。2)最近，生物科学已经能够收集关于遗传学，蛋白质和其他生物组织水平的细节的大量有价值的数据，并且现在能够前所未有地接近更广泛的研究问题，这些系统是如何组织的，以及它们的功能如何整合到一个生命整体中。这需要新的建模和分析技术，解决这些系统的复杂动态特性。在这两个学科中，这些挑战的基础是一套共同的原则和概念，这些原则和概念集中植根于复杂性科学。我们提出了一个新的复杂性科学的培训计划，解决这两个领域的核心原则和问题。我们将提供一个新的教学大纲，使用生物学各个层次的例子：从分子，细胞，组织，生物体，人口和社会群体;并将这些与示例技术系统及其挑战和需求进行比较。我们使用这些例子来教授复杂性科学的思维框架，提供超越传统工具箱所必需的建模，分析和“系统思维”。这样的例子迫使新的思维方式的设计，控制和分析：在某些情况下，生物类比已经导致建立生物启发的计算方法（神经网络，进化计算，蚁群优化），在其他情况下，新的模型和形式化提供了新的类比思考信息和通信技术系统。在另一个方向，物理科学的工具和方法，生物系统的应用提供了一个工具包的建模，形式化和抽象的复杂性科学的主题，生物科学需要向前迈进。课程将提供生物和技术的例子与建模技术和实用的分析工具，可以应用到这些领域相匹配的混合物。每天下午将提供实际的实验室会议，与会者可以使用计算模型和工具来探索和应用他们所学到的示例系统。晚上将包括小组问题练习，学生可以在轻松的气氛中进行小组练习（“酒吧练习”），以讨论和加强想法，并了解彼此的研究领域以及如何涉及复杂问题。我们将提供复杂性科学的四个不同的子主题，每个子主题都有不同的重点和例子，并教授不同的工具和分析技术：网络和鲁棒性，适应性和可塑性，进化性和可扩展性，以及沟通和集体组织。学生可以只参加一个或几个分主题。将重复两个次主题，使总共提供六个课程。

We propose to run six one-week taught courses in complexity science. 'Complexity' is used here in a special sense to refer to systems that have a large number of relatively simple parts that interact to produce collective behaviour (which may not be evident from the behaviour of the individual parts). Although historically complexity science has addressed systems as uniform as the interaction of grains in piles of sand, the richest examples of such systems are biological systems: e.g. organisms composed of cells, or metabolic systems composed of networks of interacting organic molecules. Interesting examples in man-made systems include the somewhat self-organised properties of world-wide-web content, or urban development patterns. The theme of these courses will be the unification of complexity principles from computing/information technologies and the biological sciences. This will address the new challenges that these disciplines are facing: 1) Modern Information and Communication Technology (ICT) systems need to be flexible, robust, adaptable and scalable but conventional approaches to design, control and software engineering are increasingly unable to provide this. 2) Very recently, the biological sciences have been able to collect enormous amounts of valuable data about the details of genetics, proteins, and other levels of biological organisation and it is now, like never before, able to approach broader research questions about how these systems are organised and how their functions integrate into a living whole. This requires new modelling and analysis techniques that address the complex dynamic properties of these systems. Underlying these challenges in both disciplines are a common set of principles and concepts rooted centrally in complexity science. We propose a new training programme in complexity science that tackles the core principles and issues unifying these two domains. We will provide a new syllabus using examples from all levels of biology: from molecules, to cells, to tissues, to organisms, to populations and social groups; and compare these with example technological systems and their challenges and needs. We use these examples to teach a complexity science framework of thinking that provides the modelling, analysis and 'systems mind set' necessary to move beyond conventional toolkits. Such examples force new ways of thinking about design, control and analysis: in some cases biological analogies have already led to established biologically-inspired computational methods (neural networks, evolutionary computation, ant colony optimisation), in other cases, new models and formalisms provide new analogies for thinking about ICT systems. In the other direction, the application of physical sciences tools and methods to biological systems provides a toolkit of modelling, formalisation and abstraction for the complexity science topics that the biological sciences require to move forward.The courses will provide a mixture of biological and technological examples matched with modelling techniques and practical analysis tools that can be applied to these domains. Each afternoon will provide practical lab sessions where attendees can use computational models and tools to explore and apply what they have learned to example systems. The evenings will include group problem exercises that students can do in groups in a casual atmosphere ('bar exercises') to discuss and reinforce ideas and learn about each others research areas and how complexity issues are involved. We will offer four different sub-themes on complexity science each having a different emphasis and examples, and teaching different tools and analysis techniques: Networks and Robustness, Adaptation and Plasticity, Evolvability and Scalability, and Communication and Collective Organisation. Students may attend just one sub-theme or several. Two sub-themes will be repeated making six courses offered in total.