论文标题
部分可观测时空混沌系统的无模型预测
Binding of Curvature-Inducing Proteins onto Biomembranes
论文作者
论文摘要
储层计算是预测湍流的有力工具，其简单的架构具有处理大型系统的计算效率。然而，其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间，然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用，该系统模拟了湍流的若干特征。我们表明，使用径向基函数作为非线性投影器，即使只有部分观测并且不知道控制方程，也能稳健地捕捉复杂的系统非线性。最后，我们表明，当测量稀疏、不完整且带有噪声，甚至控制方程变得不准确时，我们的网络仍然可以产生相当准确的预测，从而为实际湍流系统的无模型预测铺平了道路。
We review the theoretical analyses and simulations of the interactions between curvature-inducing proteins and biomembranes. Laterally isotropic proteins induce spherical budding, whereas anisotropic proteins, such as Bin/Amphiphysin/Rvs (BAR) superfamily proteins, induce tabulation. Both types of proteins can sense the membrane curvature. We describe the theoretical analyses of various transitions of protein binding accompanied by a change in various properties, such as the number of buds, the radius of membrane tubes, and the nematic order of anisotropic proteins. Moreover, we explain the membrane-mediated interactions and protein assembly. Many types of membrane shape transformations (spontaneous tubulation, formation of polyhedral vesicles, polygonal tubes, periodic bumps, and network structures, etc.) have been demonstrated by coarse-grained simulations. Furthermore, traveling waves and Turing patterns under the coupling of reaction-diffusion dynamics and membrane deformation are described.
