Stochastic Simulation Of Excitation-contraction Coupling

工作总结由于与Clara Franzini-Armstrong的广泛合作，我们已经获得了关于兔窦房结细胞细胞器和兰尼定受体分布的广泛统计数据。这些数据表明，我们的3D随机SANC模型的参数需要进行广泛的修改。然而，EM数据不足以确定兰尼定受体在细胞表面的临界分布。我们已经使用超分辨率SIM显微镜进行了广泛的成像，并开发了能够三维重建Ryanodine受体簇的位置和大小的软件，这将直接用于模型中。我们目前正在开发软件，该软件将使用1000个处理器来模拟单个细胞的形状和RyR分布。我们已经开发了软件，可以在模拟和实验记录中检测、分类和跟踪3D+时间的钙释放事件。这导致了对作为肾上腺素能刺激的函数在模型中发生传播的方式的新理解，并发现实验记录中的释放事件比之前怀疑的要多得多。我们已经开始研究窦房结内细胞的异质性，无论是在分离的细胞中，还是在小鼠整个窦房结标本的高空间和时间分辨率图像中。我们已经将3D随机模型扩展到多个相互作用的细胞。在接下来的程序阶段，我们将尝试模拟不同种类的相互作用的细胞产生心律的方式，作为一种紧急特性。我们还启动了一项新的研究，将固体理论中的统计物理方法应用于兰尼定受体簇集的相互作用。这表明，EC耦合过程涉及两个不同的相变，可以用解析的方式建模。根据科学顾问委员会的建议，我们正致力于将我们广泛的建模软件转换为可供其他研究人员使用的形式。这很复杂，因为模型软件是用计算机代数语言Macsyma编写的，该语言的商业形式已不再可用。为了解决这个问题，我们与该语言的一些原始开发者合作，进行了一个为期4个月的项目，以升级免费的开放源码版本(出于版权原因，称为Maxima)，以便它可以处理我们的建模软件。此升级现已包含在最新版本的Maxima(Sourceforge.com)中，因此我们的建模套件现在可以发布并供其他人使用。然而，在过去的6个月里，这项工作因新冠肺炎大流行而中断。利用流行病传播的数学结构与我们数十年来一直研究的钙诱导的钙释放几乎相同的数学结构，我们重新调整了数学和计算工具的用途，开发了一个新冠肺炎流行病学模型。这项工作提出了一种假设，即实时观察到的复杂的传播模式可以用许多预测模型中通常没有考虑的社会异质性的影响来解释。我们已经开发了一个通用的流行病模拟工具，它可以用大量的子种群来表示一个种群，其中一些子种群代表社会而不是地理单元，并通过用户指定的交互网络进行连接。这个模型是完全随机的，并包含了超级扩散的影响。我们发现，无法在社会上保持距离的弱势群体可能会推动这一流行病，并导致复杂和不可预测的传播模式，这实际上可能会在主要人口中挫败社会距离的努力。这部作品已提交供出版，并在MedRxiv上预印出版。

SUMMARY OF WORK As a result of extensive collaboration with Clara Franzini-Armstrong we have obtained extensive statistical data on the distribution of organelles and ryanodine receptors in rabbit siono-atrial node cells. These data indicated that the parameters of our 3D stochastic SANC model need to be extensively revised. However, the EM data are not sufficient to define the critical distribution of ryanodine receptors on the cell surface. We have done extensive imaging using ultra-resolution SIM microscopy, and have developed software that enables 3D reconstruction of the location and size of ryanodine receptor clusters, which will be used directly in the model. We are currently developing software that will use 1000 processors to model the shape and Ryr distribution of individual cells. We have developed software that can detect, classify and track calcium release event in 3D+time, both in simulations and in experimental records. This has led to new understanding of the way that propagation occurs in the model as a function of adrenergic stimulation, and to the discovery that there are many more release events in experimental records than previously suspected. We have begun studies of heterogeneity of cells within the sinus node, both in isolated cells and in high space and time resolution images of whole sinus node preparations from mouse. We have extended the 3D stochastic model to multiple, interacting cells. In the next program period we will attempt to model the way that heterogeneous interacting cells give rise to the heart rhythm as an emergent property. We have also initiated a new study that applies statistical physics methods from solid-state theory to the interactions of clustered ryanodine receptors. This has shown that the process of EC coupling involves two different phase transitions that can be modeled analytically. As advised by the Board of Scientific Counselors, we are undertaking to translate our extensive modeling software into a form that can be used by other investigators. This has been complicated because the model software in written in the computer algebra language Macsyma whose commercial form is no longer available. To solve this problem we undertook a 4 month project in collaboration with some of the original developers of the language to upgrade the free, open-source version (called Maxima for copyright reasons) so that it can process our modeling software. This upgrade has now been incorporated in the latest version of Maxima (Sourceforge.com) so that our modeling suite can now be published and used by others. However, during the past 6 months this work has been interrupted by the Covid-19 pandemic. Taking advantage of the fact that spread of an epidemic has a mathematical structure almost identical to that of the calcium-induced calcium release we have been studying for decades, we have repurposed our mathematical and computational tools to develop a model of Covid-19 epidemiology. This work addressed the hypothesis that the complex patterns of propagation being observed in real time could be explained by the effect of social heterogeneity not generally accounted for in many forecasting models. We have developed a general epidemic simulation tool that can represent a population by a large number of sub-populations, some representing social rather than geographic units, and connected by a network of interactions specified by the user. This model is fully stochastic and incorporates the effects of super-spreading. We find that vulnerable subgroups unable to socially-distance can drive the epidemic and lead to complex and unpredictable patterns of spread that can actualy defeat the efforts of social distancing in the main population. This work has been submitted for publication and published in a pre-print on MedRxiv.