Learning the rules of collective cell migration using deep attention networks.
Learning the rules of collective cell migration using deep attention networks.
复制标题
DOI：
10.1371/journal.pcbi.1009293
复制
发表时间：
2022-04
影响因子：
4.3
通讯作者：
中科院分区：
文献类型：
--
作者：
集体，协调的细胞运动支撑着所有多细胞生物的关键过程，但它一直难以同时表达这些运动背后的“规则”，清晰，可解释的形式，有效地捕捉高维细胞间的相互作用动力学的方式是直观的研究人员。在这里，我们应用深度注意力网络来分析几种典型的活组织系统，并仅使用细胞迁移轨迹数据来呈现每种组织类型的潜在集体迁移规则。我们使用这些网络来学习具有不同集体行为的关键组织类型的行为-上皮，内皮和转移性乳腺癌细胞-并展示结果如何补充传统的生物物理方法。特别是，我们提出了注意力地图，指示相邻细胞的相对影响，学习转向决定的“焦点细胞”-在集体设置感兴趣的主要细胞。通俗地说，我们把这种学习到的相对影响称为“注意力”，因为它作为物理参数的代理，修改焦点细胞的未来运动作为每个相邻细胞的函数。这些注意力网络揭示了每个模型组织独特的影响和注意力的不同模式。内皮细胞表现出密切关注其最直接的最前沿的邻居，而在更膨胀的上皮组织中的细胞更广泛地受到邻居在一个相对较大的前向扇区的影响。更多间充质细胞、转移性细胞的集合的注意力地图揭示了完全对称的注意力模式，表明缺乏任何特定的协调或感兴趣的方向。此外，我们还展示了注意力网络如何能够检测和学习这些规则如何根据生物物理背景（如组织内的位置和细胞拥挤）发生变化。这些结果只需要细胞轨迹，没有建模假设，突出了注意力网络的潜力，为复杂的细胞系统提供进一步的生物学见解。集体行为对多细胞生命的功能至关重要，大规模协调的细胞迁移使器官形成过程能够协调皮肤愈合。然而，我们缺乏有效的工具来发现和清晰地表达单个细胞级别的集体规则。在这里，我们采用了一个精心构造的神经网络，直接从细胞轨迹数据中提取集体信息。该网络基于来自各种系统的数据进行训练，包括显示视觉上不同形式的集体运动的典型集体细胞系统（HUVEC和MDCK细胞），以及高度不协调的转移性癌细胞（MDA-MB-231）。使用这些经过训练的网络，我们可以为每个系统生成注意力地图，它指示组织内的细胞如何从周围的邻居那里获取信息，作为分配给这些邻居的权重的函数。因此，对于其中细胞倾向于遵循前面细胞的路径的细胞类型，注意力图将显示焦点细胞空间前方的细胞的高权重。我们目前的结果在额外的指标，如准确性图和相互作用的细胞的数量，并鼓励未来的发展改进的指标。
Collective, coordinated cellular motions underpin key processes in all multicellular organisms, yet it has been difficult to simultaneously express the ‘rules’ behind these motions in clear, interpretable forms that effectively capture high-dimensional cell-cell interaction dynamics in a manner that is intuitive to the researcher. Here we apply deep attention networks to analyze several canonical living tissues systems and present the underlying collective migration rules for each tissue type using only cell migration trajectory data. We use these networks to learn the behaviors of key tissue types with distinct collective behaviors—epithelial, endothelial, and metastatic breast cancer cells—and show how the results complement traditional biophysical approaches. In particular, we present attention maps indicating the relative influence of neighboring cells to the learned turning decisions of a ‘focal cell’–the primary cell of interest in a collective setting. Colloquially, we refer to this learned relative influence as ‘attention’, as it serves as a proxy for the physical parameters modifying the focal cell’s future motion as a function of each neighbor cell. These attention networks reveal distinct patterns of influence and attention unique to each model tissue. Endothelial cells exhibit tightly focused attention on their immediate forward-most neighbors, while cells in more expansile epithelial tissues are more broadly influenced by neighbors in a relatively large forward sector. Attention maps of ensembles of more mesenchymal, metastatic cells reveal completely symmetric attention patterns, indicating the lack of any particular coordination or direction of interest. Moreover, we show how attention networks are capable of detecting and learning how these rules change based on biophysical context, such as location within the tissue and cellular crowding. That these results require only cellular trajectories and no modeling assumptions highlights the potential of attention networks for providing further biological insights into complex cellular systems. Collective behaviors are crucial to the function of multicellular life, with large-scale, coordinated cell migration enabling processes spanning organ formation to coordinated skin healing. However, we lack effective tools to discover and cleanly express collective rules at the level of an individual cell. Here, we employ a carefully structured neural network to extract collective information directly from cell trajectory data. The network is trained on data from various systems, including canonical collective cell systems (HUVEC and MDCK cells) which display visually distinct forms of collective motion, and metastatic cancer cells (MDA-MB-231) which are highly uncoordinated. Using these trained networks, we can produce attention maps for each system, which indicate how a cell within a tissue takes in information from its surrounding neighbors, as a function of weights assigned to those neighbors. Thus for a cell type in which cells tend to follow the path of the cell in front, the attention maps will display high weights for cells spatially forward of the focal cell. We present results in terms of additional metrics, such as accuracy plots and number of interacting cells, and encourage future development of improved metrics.