Dissemination of the Human Neocortical Neurosolver (HNN) software for circuit level interpretation of human MEG/EEG

HNN U24传播项目摘要 人类大脑皮质神经解算器(HNN)神经建模工具是在Brain Initiative的资助下开发的 (R01EB022889：09/2016-06/2020)以实现该倡议的目标，即“开发创新技术以了解 大脑回路和回路集合，为理解人脑和机制提供信息 治疗它的紊乱“。HNN是一种生物物理原理的新皮质电路模型，具有适当的物理特性 允许从大尺度的人类磁图和脑电(M/EEG)信号连接到其细胞 和电路级生成器。HNN是一个假设开发和测试工具，其设计和能力 与其他M/EEG神经建模软件相比是独一无二的。HNN的一个关键价值是在功能上连接- 在动物模型中研究的与电路水平动力学相关的人体信号，包括来自革命性的数据 用于小鼠和猴子的遗传和成像工具(例如，神经像素记录、钙和电压成像)。 这些联系对于发现大脑信息处理的新原理和开发 当这一过程被神经病理扰乱时进行治疗。M/EEG被广泛使用，而且有无数种 从这些关于人脑功能和健康的信号中得出的推论：HNN提供了一个高度可访问的 为研究人员提供了一种工具，用于与这些信号背后的详细神经元和电路进行原则性连接， 在电路级现实中测试新想法和基础结论。 神经科学界积极参与将HNN用于基础和临床研究，包括 研究阿尔茨海默氏症、自闭症谱系障碍、疼痛、抑郁和健康发展。而当 尽管成功开发了HNN，但在增长和长期可持续性方面仍存在一些挑战。这个 这项提议的目标是支持HNN的传播，以供广泛使用和社区驱动 发展。确定了维护HNN代码和文档方面的挑战，以确保可查找性， 将讨论可访问性、互操作性和可重用性(目标I)，并对以下方面进行必要的关键增强 将为最终用户需求提供支持，并对模型得出的预测进行实验验证(目标二)。 用HNN的大规模神经网络解释M/EEG信号复杂的多尺度起源往往需要领域 在计算神经建模、人类M/EEG和神经动力学方面的专长。为了满足这一需求，我们将 继续与社区合作，通过研讨会、直接协作、 并与最终用户组协商，并在可自由访问的超级计算机上启用HNN模拟 (目标III)。最终用户反馈和记录的支持需求将用于制定可持续发展计划。一个 世界级指导委员会包括广泛采用的神经科学软件的开发者，他们将分享 他们的专业知识帮助HNN发挥其作为翻译神经科学工具的最大潜力。

HNN U24 DISSEMINATION PROJECT SUMMARY The Human Neocortical Neurosolver (HNN) neural modeling tool was developed with BRAIN Initiative funding (R01EB022889: 09/2016–06/2020) to meet the Initiative’s goal to “develop innovative technologies to understand brain circuits and ensembles of circuits that inform understanding of the human brain and mechanisms for treating its disorder”. HNN is a biophysically principled neocortical circuit model with appropriate physics that allow bridging from macroscale human magneto- and electro-encephalography (M/EEG) signals to their cellular and circuit level generators. HNN is a hypothesis development and testing tool whose design and capabilities are unique compared to other M/EEG neural modeling software. A key value of HNN is to connect functionally- relevant human signals to circuit level dynamics studied in animal models, including data from revolutionary genetic and imaging tools used in mice and monkeys (e.g., Neuropixel recordings, calcium, and voltage imaging). These links are essential to discovering new principles of brain information processing and to developing treatments when this processing is disrupted by neuropathology. There is widespread use of M/EEG, and myriad inferences drawn from these signals about human brain function and health: HNN provides a highly accessible tool for researchers to make principled connections to the detailed neurons and circuits underlying these signals, to test new ideas and ground conclusions in circuit-level reality. The neuroscience community is actively engaged in the use of HNN for basic and clinical research, including studies of Alzheimer’s disease, autism spectrum disorder, pain, depression, and healthy development. While HNN was successfully developed, there remain several challenges for growth and long-term sustainability. The goal of this proposal is to support dissemination of HNN for broadly accessibly use and community-driven development. Identified challenges in maintenance of HNN’s code and documentation that ensure Findability, Accessibility, Interoperability, and Reusability will be addressed (Aim I), and key enhancements necessary to support end-user needs and experimental validation of model-derived predictions will be developed (Aim II). Interpreting the complex multiscale origin of M/EEG signals with HNN’s large-scale neural often requires domain expertise in computational neural modeling, human M/EEG, and neural dynamics. To support this need, we will continue to work with the community to integrate HNN into their projects through workshops, direct collaboration, and consultation with end-user groups, and by enabling HNN simulation on freely accessible supercomputers (Aim III). End-user feedback and documented support needs will be used to develop a Plan for Sustainability. A world-class Steering Committee includes developers of widely-adopted neuroscience software who will share their expertise to help HNN reach its maximal potential as translational neuroscience tool.