The final frontier in connectomics: Forward engineering brain networks
The final frontier in connectomics: Forward engineering brain networks
复制标题
连接组学的最后前沿：正向工程大脑网络
DOI：
10.1016/j.plrev.2019.11.004
复制
发表时间：
2019
影响因子：
11.7
通讯作者：
Towlson, Emma K.
中科院分区：
文献类型：
--
作者：
Towlson, Emma K.
它读起来像是科幻小说。在这个世界里，我们可以设计大脑中神经元之间的任意连接模式，并在生命系统中实现它们。在这篇综述中，拉比诺维奇提出，这个世界可能比我们想象的要近[1]。合成连接的可能性有哪些？我们该如何着手设计它们呢？它的社会影响和技术应用又是什么？在许多新的神经技术的出现和跨机构和国家的大规模协调努力的推动下，连接学进入了爆炸性发展的时代。合成神经生物学正在迅速发展，并为连接组图[2]提供了重要贡献，例如通过基因测序[3]、病毒追踪[4]和荧光成像技术[5]。理论家和数据分析师正在竞相跟上多模式数据的数量和新颖性，这些数据具有前所未有的细节[6]。网络神经科学已经取得了长足的进步，确定了跨物种连接的共同特征，如丰富的俱乐部[7-9]，为连接尺度和数据类型[10，11]奠定了基础，并揭示了接线图、动力学和行为之间的关系[12，13]。然而，其中一些最重要和最顽固的问题，包括从分子、神经元到区域的水平如何相互联系和相互作用，以及对生物行为从神经元活动中出现的方式的理解，仍然只有部分答案。合成连接体的前景是神经科学家终极实验场地的前景。一个向前推进工程模型系统的机会。最先进的模拟，如OpenWorm[14]和虚拟大脑[15]，已经提供了一个令人印象深刻的计算环境，可以在其中研究一些神经系统的动力学，但合成连接体在这方面走得更远。我们越来越多地认识到，不能孤立地理解大脑。也就是说，它作为一个更大的系统--身体--的一个组成部分而存在，这个系统与环境相互作用。我们知道，生物体的行为根据它们的身体状态而不同，例如线虫以一种随饱腹感而变化的方式导航化学感觉梯度[16]。一个合成的连接体巧妙地绕过了建模
It reads like science fiction. A world in which we can design arbitrary wiring patterns between the neurons in our brains, and realise them in living systems. In this review, Rabinowitch posits that this world may be closer than we think [1]. What are the possibilities synthetic connectomes will enable? How might we go about engineering them? And what are the societal implications and technological applications? Connectomics has entered an era of explosive progression, driven by the advent of many novel neurotechnologies and large-scale coordinated efforts across institutions and countries. Synthetic neurobiology is rapidly progressing and providing important contributions to connectome mapping [2], for example through genetic sequencing [3], viral tracing [4], and fluorescence imaging techniques [5]. Theoreticians and data analysts are racing to keep up with the volume and novelty of multimodal data of unprecedented detail [6]. Network neuroscience has made strides, identifying common features in connectomes across species such as rich clubs [7–9], laying the groundwork for connecting scales and data types [10, 11], and shedding light on the relationships between the wiring diagram, dynamics, and behaviours [12, 13].Yet, some of these most important and stubborn questions, including how levels–from the molecular, to the neuronal, to the regional–relate to and interact with each other, and an understanding the ways in which organism behaviour emerges from neuronal activity, remain only partially answered. The promise of a synthetic connectome is the promise of the ultimate experimental playground for neuroscientists. An opportunity to forward engineer model systems. State of the art simulations such as those from OpenWorm [14] and the virtual brain [15] already offer an impressive computational environment in which to investigate dynamics for some neural systems, but the synthetic connectome takes this so much further. Increasingly we are acknowledging that the brain cannot be understood in isolation. That is to say, it exists as a component of a larger system–the body–which interacts with an environment. We know that organisms behave differently according to their bodily state, for instance C. elegans navigates chemosensory gradients in a way that varies with satiety [16]. A synthetic connectome neatly bypasses modelling