Interactive Modeling of Multi-scale Biological Systems

我对理解看似复杂的系统的涌现特性很感兴趣，在这些系统中，“总和大于部分”。十多年来，我一直与生物学家、医学研究人员和医学教育工作者合作，设计和测试生物系统的计算模型。生物学家倾向于认为他们的研究对象是离散的实体。在计算上，这些实体可以用“代理”来表示，这些代理的交互规则是已知的或可以识别的，最终用带有简明算法描述的“简单程序”来表示。一旦这些相互作用的元素被安排在隔间内，这些隔间又被组织成相互连接的更大的单位（“宇宙”），复杂的行为就出现了。作为一名计算机科学家，我研究如何通过三维交互式计算机模型来捕捉、说明并使看似复杂的系统变得可探索。我们在进化与群体设计研究实验室开发的基于主体的模型，有助于传达我们对复杂系统的基本理解。除了精心制作的动画，开发新的算法和方法来构建这样的模型不仅加强了“系统”科学的研究，而且为在模拟3D空间中生成、可视化和探索复杂模型创造了新的计算方法。“看到了才开始明白”。通过互动模式探索自己的学习路径，加深了我们的理解。我接下来的研究重点是如何在空间和时间上整合不同分辨率水平的多尺度模型。我目前的研究重点是如何将连续数学模型与基于智能体的方法相结合。数学模型捕获大量粒子（如分子）的平均浓度变化，而代理方法更适合表示少量出现的实体，并且这些实体（如蛋白质或细胞）之间的特定相互作用可以在3D中建模到高水平的细节。由于许多复杂系统是由一系列尺度上的子系统组成的（例如，从系统和器官到组织、细胞和蛋白质），多尺度计算模型需要包含和整合发生在不同空间尺度和不同时间分辨率上的过程。使多尺度模型易于访问是我研究的另一个重要方面。我研究用户界面的可视化，导航，表示，并跨越时间和空间尺度的模型层次结构的探索。同时，我希望我们的模型能够实时运行，同时在任何时候都保持交互性。通过探索在移动计算平台上运行模拟的解决方案，我努力使更广泛的用户群能够访问复杂的模型。

I am interested in understanding in emergent properties of seemingly complex systems, where “the sum is greater than its parts”. For more than a decade, I have been collaborating with biologists, medical researchers, and medical educators to design and test computational models of biological systems. Biologists tend to think about their objects of study as discrete entities. Computationally, these entities can be represented by ‘agents’, whose interaction rules are known or can be identified, and ultimately represented by ‘simple programs’ with concise algorithmic descriptions. Once such interacting elements are arranged within compartments, which are organized into interconnected larger units (‘universes’), complex behaviors emerge. As a computer scientist I look at how to capture, illustrate and make seemingly complex systems explorable through 3-dimensional, interactive computer models. The agent-based models, that we develop in my Evolutionary & Swarm Design research lab, help to communicate our fundamental understanding of complex systems. Going beyond carefully crafted animations, the development of new algorithms and methodologies to build such models does not only enhance research in “systems” sciences but creates new computational approaches for generating, visualizing, and exploring complex models in simulated 3D spaces. “To see is to begin to understand”. And to explore our own learning paths through interactive models deepens our understanding. The next steps in my research focus on how to integrate multi-scale models across different levels of resolution in space and time. My current research focus is on how to combine continuous mathematical models with agent-based approaches. The mathematical models capture concentration changes averaged over a large number of particles (such as molecules), whereas agent methods are a better fit for representing entities that appear in small quantities, and where specific interactions among these entities (such as proteins or cells) can be modeled to a high level of detail in 3D. As many complex systems are composed of a hierarchy of subsystems across a range of scales (e.g., from systems and organs to tissues, cells, and proteins), multi-scale computational models need to incorporate and integrate processes that occur at different spatial scales and at different resolutions in time. Making multi-scale models accessible is another important aspect in my research. I investigate user interfaces for visualization, navigation, presentation, and exploration of model hierarchies across temporal and spatial scales. At the same time, I want our models to be run in real time while maintaining interactivity at any point. I strive to make complex models accessible to a wider user base, by exploring solutions for running simulations on mobile computing platforms.