Knowledge “installed” diffusion model predicts the geometry of actin cytoskeleton from cell morphology
Knowledge “installed” diffusion model predicts the geometry of actin cytoskeleton from cell morphology
复制标题
DOI：
10.1101/2023.01.12.523863
复制
发表时间：
2023-01
期刊：
影响因子：
--
通讯作者：
Honghan Li;Shiyou Liu;S. Deguchi;D. Matsunaga
中科院分区：
文献类型：
--
作者：
Honghan Li;Shiyou Liu;S. Deguchi;D. Matsunaga
细胞由于其生理活动而表现出各种形态特征，并且细胞形态的变化固有地伴随着肌动蛋白细胞骨架的组装和拆卸。应力纤维是基于肌动蛋白的细胞内结构的重要组成部分，并且高度参与许多生理过程，例如，机械转导和维持细胞形态。虽然它被广泛接受，细胞几何形状的变化与应力纤维的分布和定位相互作用，它仍然不清楚，如果有潜在的几何原则之间的细胞形态和肌动蛋白细胞骨架。在这里，我们提出了一个机器学习系统，它使用扩散模型，可以将细胞形状转换为应力纤维的分布。通过使用细胞形状和应力纤维的相应数据集进行训练，我们的系统学习转换以从其相应的细胞形状生成应力纤维图像。预测的应力纤维分布与实验数据吻合较好，预测的应力纤维与实验观察到的应力纤维重叠区域达到79.3 ± 12.4%。我们发现了一些未知的性质，如应力纤维长度和细胞面积之间的线性关系。通过细胞形态和相应应力纤维定位之间的这种“安装”转换关系，我们的系统可以执行虚拟实验，该虚拟实验提供了显示虚拟细胞形状的应力纤维分布概率的可视化地图。我们的系统提供了一个强大的方法，以寻求进一步隐藏的几何原则之间的细胞形态和肌动蛋白细胞骨架。
Cells exhibit various morphological characteristics due to their physiological activities, and changes in cell morphology are inherently accompanied by the assembly and disassembly of the actin cytoskeleton. Stress fibers are a prominent component of the actin-based intracellular structure and are highly involved in numerous physiological processes, e.g., mechanotransduction and maintenance of cell morphology. Although it is widely accepted that variations in cell geometry interact with the distribution and localization of stress fibers, it remains unclear if there are underlying geometric principles between the cell morphology and actin cytoskeleton. Here we present a machine learning system, which uses the diffusion model, that can convert the cell shape to the distribution of stress fibers. By training with corresponding datasets of cell shape and stress fibers, our system learns the conversion to generate the stress fiber images from its corresponding cell shape. The predicted stress fiber distribution has good agreement with the experimental data, and the overlap region of predicted and experimentally observed stress fibers reaches 79.3 ±12.4%. We then found some unknown natures such as a linear relation relationship between the stress fiber length and cell area. With this “installed” conversion relation between cellular morphology and corresponding stress fibers’ localization, our system could perform virtual experiments that provide a visual map showing the probability of stress fiber distribution from the virtual cell shape. Our system provides a powerful approach to seek further hidden geometric principles between the cell morphologies and actin cytoskeletons.