Neuroinformatics Core

项目摘要 将在堪萨斯州立大学（KSU）建立一个神经信息学（NI）核心， 跨C-NAP的项目、计划和核心。NI核心将提供安全、快速和高效的访问， 数据共享、分析、计算建模和高性能计算（HPC）。该芯 对于支持管理、分析和建模收集的大型数据集尤为重要， 神经科学研究，如神经成像（fMRI，EEG，扩散张量成像），眼动跟踪，驾驶 模拟器行为和其他经常实时进行的定量行为测量。NI 核心访问将提供给C-NAP的所有主要成员和附属成员。NI核心将包括 Web服务器和双数据服务器（最初每个服务器的原始容量约为100 TB），以提供冗余， 可扩展性，计划每年购买更多存储服务器（或升级初始存储服务器）， 随着C-NAP需求的增长，预计将达到1 PB的总需求。这些服务器将与高速 （最低10 GbE）网络连接到KSU和WSU校园，40 GbE连接到Beocat群集， 堪萨斯州最大的超级计算机集群。NI核心将用于长期存档， 数据共享，以满足NSF/NIH的要求，由经过身份验证的用户通过自己的云（或 等效）存储系统。这将促进合作研究的发展和延续 通过为数据存储和用户组访问提供安全的空间，核心将被引导 丹尼尔安德森博士，他是KSU工程计算研究所所长 和科学，在高性能计算领域的专家和极端的科学和工程 发现环境（XSEDE）3级服务提供商。技术支持将由全职 数据技术员和研究生。国家机构核心的首要目标是提高国家行动方案的能力， 研究人员通过整合解决这些问题所需的技术来竞争校外资金， 这是现代神经科学研究者面临的最具挑战性的问题。

PROJECT SUMMARY A Neuroinformatics (NI) core will be established at Kansas State University (KSU) to support the primary projects, programs, and cores across C-NAP. The NI core will provide secure, fast, and efficient access for data sharing, analytics, computational modeling, and high-performance computing (HPC). This core is particularly important for supporting the management, analysis, and modeling of large data sets collected in neuroscience studies such as neuroimaging (fMRI, EEG, diffusion tensor imaging), eye tracking, driving simulator behavior, and other quantitative behavioral measurements that are often conducted in real-time. NI core access will be available to all primary and affiliated members of C-NAP. The NI core will be comprised of web servers, and dual data servers (initially each specified at ~100TB raw capacity) to provide redundancy and scalability, with plans to purchase additional storage servers (or upgrade the initial ones) each following year to an expected 1PB aggregate total as C-NAP needs grow. These servers will be linked with high-speed (minimum 10GbE) networks to the KSU and WSU campuses and at 40GbE to the Beocat cluster, one of the largest supercomputing clusters in the state of Kansas. The NI core will be used for longer-term archiving and data sharing, to meet NSF/NIH requirements, with access by authenticated users through the Own Cloud (or equivalent) storage system. This will promote the development and continuation of collaborative research activities by providing a secure space for data storage and access by groups of users. The core will be directed by Dr. Daniel Andresen, who is the Director of the KSU Institute for Computational Research in Engineering and Science, an expert in the area of high performance computing and an Extreme Science and Engineering Discovery Environment (XSEDE) Level 3 Service Provider. Technical support will be provided by a full-time data technician and graduate student. The overarching goal of the NI core is to promote the ability of C-NAP researchers to compete for extramural funding by incorporating technologies that are needed to answer the most challenging questions facing modern neuroscience researchers.