CAREER: Mesoscale computational modeling of intracellular soft matter

细胞内的缠结或交联聚合物网络，如肌动蛋白和微管细胞骨架，与来自环境的机械信号相互作用并对其做出反应，影响细胞的生长和分裂、干细胞如何分化为特定的细胞类型、以及癌细胞如何增殖等重要的分子和细胞现象。在许多情况下，这些网络结构是通过脂膜和细胞质流动的机械反馈而动态改变的。然而，目前的大多数建模和计算方法都集中在计算的可处理性上，而忽视了这些机械反馈。该项目的一个中心主题是，开发细胞内聚合物网络(和其他生物物质)的物理现实模型需要一种计算方法，该方法有效地将网络的力学与细胞内流体流动的流体动力学以及与生物膜的机械相互作用结合在一起。这种综合的计算努力是更好地理解细胞内现象的关键。特别是，它们有可能帮助诊断和治疗细胞骨架相关的疾病，并指导干细胞技术的发展。拟议的工作将有助于研究生和本科生的科学和专业发展。它还将整合极好的教育机会，包括开发开放式教育资源；通过开发与拟议项目有关的专门计算机编码活动，促进计算机编码扫盲，以此作为研究物理现象的手段；举办训练营，培训本科生“细胞内软物质的中尺度计算模型”；与布朗大学科学中心合作，吸引来自代表性不足社区的高中生参与科学计算和软物质物理学。DPD可以准确地模拟聚合物和流体的分子尺度运动，而不需要计算负担，这通常会使替代模拟(例如分子动力学)变得不可行。另一方面，随机均匀化方法提供了从中尺度DPD描述中有效地提取整体力学性质的手段。该项目的具体目标包括：(1)开发专门的基于粒子的方法，用于研究细胞质流动和细胞内物质之间的机械相互作用；(2)对随机均质化方法进行计算和分析研究，以开发细胞内物质的介观力学描述；(3)将所开发的算法应用于生物物质的具体实例的计算建模，例如肌动蛋白细胞骨架和DNA大分子；以及(4)模拟数据的聚集和分类，以表征所研究系统中生物物质的动力学。考虑到最近对细胞内聚合物网络(例如肌动蛋白细胞骨架)作为治疗靶点的兴趣，包括癌症、神经退行性疾病和肾脏疾病，所提出的计算方法有望促进重要的生物医学应用，特别是细胞骨架动力学的治疗性扰动。

Intracellular networks of entangled or cross-linked polymers, such as the actin and microtubule cytoskeleton, interact with and respond to mechanical cues from the environment, influencing how cells grow and divide, how stem cells differentiate into specific cell types, and how cancer cells proliferate, among other important molecular and cellular phenomena. In many cases, the structure of these networks is dynamically altered by the mechanical feedback of lipid membranes and cytoplasmic flows. However, the majority of current modeling and computational approaches focus on computational tractability at the expense of these mechanical feedbacks. A central theme of this project is that the development of physically realistic models of intracellular polymer networks (and other biological matter) requires a computational approach that efficiently integrates the mechanics of the network with the hydrodynamics of intracellular fluid flows and the mechanical interactions with biological membranes. Such integrative computational efforts are key to building a better understanding of intracellular phenomena. In particular, they hold the potential to help diagnose and treat cytoskeleton-related diseases and guide the development of stem cell techniques. The proposed work will help with the scientific and professional development of both graduate students and undergraduate students. It will also integrate excellent educational opportunities including the development of open education resources; promoting computer-coding literacy as a means to investigate physical phenomena through the development of an educational website with specialized computer coding activities pertaining to the proposed project; conducting boot camps to train undergraduate students on "Mesoscale computational modeling of intracellular soft matter" and; working with the Science Centre at Brown University to engage high school students from underrepresented communities in scientific computing and the physics of soft matter.The PI proposes a synergistic combination of approaches consisting of dissipative particle dynamics (DPD) simulations and stochastic homogenization methods. DPD can accurately model the molecular-scale motion of polymers and fluids without the computational burden that normally makes alternative simulations (e.g., molecular dynamics) unviable. Stochastic homogenization methods, on the other hand, provide the means of efficiently extracting bulk mechanical properties from the mesoscale DPD description. Specific aims of the project include (i) the development of specialized particle-based methods for investigating the mechanical interactions between cytoplasmic flows and intracellular matter, (ii) a computational and analytic investigation of stochastic homogenization approaches to developing mesoscopic mechanical descriptions of intracellular matter, (iii) applications of the developed algorithms to the computational modeling of specific examples of biological matter, such as the actin cytoskeleton and DNA macromolecules, and (iv) the clustering and classification of simulation data in order to characterize the dynamics of biological matter in the systems under investigation. Given the recent interest in intracellular polymer networks (e.g., the actin cytoskeleton) as therapeutic targets in a broad array of pathological conditions, including cancer, neurodegenerative diseases, and kidney disease, the proposed computational approach is expected to advance significant biomedical applications, especially therapeutic perturbations of cytoskeletal dynamics.