Data Science Core

双光子钙成像结合全息光遗传学激发提供了一种强大的新方法， 读取和控制神经元代码，有可能直接连接单个神经元的时空活动 解决行为。在这样的实验中产生的数据量非常大， 输入、获取的图像和行为输出。数据复杂性的另一个贡献者是全息 产生的光模式允许神经元群的光遗传学控制。既没有数据也没有分析 对于这种技术，方法是标准化的。数据科学资源核心将是必不可少的， 实现这些研究项目中提出的实验研究，这些研究项目跨越多个实验室， 感觉形态数据科学核心的三个主要目标是：（1）促进数据聚合，（2） 发展数据采集和分析的质量控制;（3）使理论指导实验设计成为可能。 第一个目标是使数据汇总标准化。我们将部署一个新的原型大脑平台图形 用户界面使用来自新兴的Neurodata Without Borders结构的数据格式。核心会起作用的 与所有团队成员一起将以多种语言编写的现有代码集成到Brain Platform中， 因此，将成为数据汇总和共享的地方枢纽。一旦实施，培训材料将 为广泛传播大脑平台而开发 第二个目标是确保数据和分析的质量。分析所有团队的代码， 通过大脑平台提供的信息将定期相互验证，并与地面实况进行对比。 在技术核心中获得的数据和其他基线数据。此外，统计分析方法将 开发和实施，以实现模型推理的功效计算。 第三个目标是使理论指导的实验设计和验证。这一目标也将支持数据 基于高维推理的分析，包括格兰杰因果关系、网络关键性和稳定性 分析和交叉点信息。 总之，数据科学核心的三个目标-标准化的数据聚合，质量控制 分析和理论指导的实验设计-将支持稳健、可靠和可重复的采集 并分析了使用双光子钙成像结合 全息光遗传刺激。

Two-photon calcium imaging combined with holographic optogenetic excitation offers a powerful new way to read and control neuronal codes, with the potential to directly link spatiotemporal activity at single-neuron resolution to behavior. The volume of data produced in such experiments is very large, involving sensory inputs, acquired images, and behavioral outputs. Another contributor to data complexity are the holographic light patterns produced that allow for optogenetic control of neuronal groups. Neither data nor analysis approaches are standardized for this technology yet. The Data Science Resource Core will be essential to enable the experimental studies proposed in these research projects, which span multiple laboratories and sensory modalities. The three key aims of the Data Science Core will be (1) to facilitate data aggregation, (2) to develop quality control on data acquisition and analysis, and (3) to enable theory guided experimental design. The first aim is to standardize aggregation of data. We will deploy a new prototype Brain Platform graphical user interface using data formats from the emerging Neurodata Without Borders structure. The Core will work with all team members to integrate existing code written in multiple languages into the Brain Platform, which will thus serve as local hubs for data aggregation and sharing. Once implemented, training materials will be developed for broad dissemination of the Brain Platform The second aim is to ensure the quality of data and analysis. Analysis codes from all team groups that is made available through the Brain Platform will be regularly validated against each other and against ground-truth data obtained in the technology core and other baseline data. Additionally, statistical analysis approaches will be developed and implemented to enable power calculations for model inference. The third aim is to enable theory guided experiment design and validation. This aim also will support data analysis based on high-dimensional inference, including Granger causality, network criticality and stability analysis, and intersection information. Together, the three aims of the Data Science Core — standardized data aggregation, quality-controlled analysis, and theory guided experimental design — will support robust, reliable and reproducible acquisition and analysis of neural and behavioral data from experiments that use two-photon calcium imaging combined with holographic optogenetic stimulation.