CAREER: Towards the Next Generation of Data-Driven Computational Brain Analytics

老道明大学被授予早期教师职业发展补助金，以支持纪水旺博士进行研究，从而更好地了解大脑。大脑是一个极其复杂的系统，因此对大脑数据的分析也是一个同样巨大的挑战。人类大脑包含数十亿个神经元和数万亿个突触(连接)；它们在基本的生物化学、功能和动力学方面都是独一无二的。大脑也是一个跨不同空间尺度组织的多层次系统，从基因、突触和细胞到电路、大脑区域和系统。如今，脑科学正在经历着日新月异的变化，并有望在不久的将来取得重大进展。最近的技术创新使科学家能够以更快的速度和更高的分辨率捕捉基因表达模式、连接性和神经元活动。这就产生了大量的数据，这些数据捕捉到了不同组织级别的大脑活动。为了解决分析这些新数据的主要挑战，该项目将开发一类高效、综合、多维、预测和相关的技术，并使用它们来分析大规模、高分辨率和多模式的脑数据集。具体地说，该项目将开发分析工具，根据基因转录图谱预测细胞分辨率、全脑连接组(“接线图”)。这一分析将阐明从基因到连接性并最终发挥作用的信息途径。该项目还将通过执行多维网络相关分析来整合其他大脑维度。此外，这个项目将解决基因表达、细胞类型和大脑结构之间的关系。该项目的成功将是一种新的高效、稳健的分析方法，这些方法足够灵活，可以整合、建模和挖掘当前和未来的大脑数据。该项目的结果将立即对多个学科产生强烈影响，包括脑数据分析和计算神经科学、生物图像信息学和大数据分析。未来的长期目标是发现大脑功能正常和受损之间的基本潜在差异。大脑数据分析的统一处理将很容易转变为培训下一代计算生物学家的新课程。该项目的多学科性质为将其组成部分纳入现有课程提供了独特的机会。脑科学已被证明是激发K-12学生科学兴趣的宝贵资源。该项目的组成部分将被整合到现有的高中生实习计划中，从而激励未来的理科学生。将特别鼓励代表人数不足的学生在整个项目中参与。该项目的成果将以同行评议出版物、开放源码软件、教程、研讨会和讲习班的形式传播。所有发现、出版物、软件和数据将在项目网站上公开提供：http://compbio.cs.odu.edu/CAREER/

Old Dominion University is awarded an Early Faculty Career Development grant to support Dr Shuiwang Ji in research leading to a better understanding of the brain. The brain is an enormously complex system, and the analysis of brain data is thus an equally enormous challenge. Human brains contain billions of neurons and trillions of synapses (junctions); and each of them is unique in their basic biochemistry, functions, and dynamics. The brain is also a multi-level system organized across different spatial scales, ranging from genes, synapses, and cells to circuits, brain regions, and systems. Today, brain science is experiencing rapid changes and is expected to achieve major advances in the near future. Recent technological innovations are enabling scientists to capture the gene expression patterns, connectivity, and neuronal activities at increasing speed and resolution. This is generating a deluge of data that capture the brain activities at different levels of organization. To attack the central challenges of analyzing these new data, this project will develop a class of efficient, integrative, multidimensional, predictive, and correlative techniques and use them to analyze large-scale, high-resolution, and multi-modality sets of brain data. Specifically, this project will develop analytics tools to predict the cellular-resolution, brain-wide connectome ("wiring diagram") from genetic transcriptional profiles. This analysis will elucidate the information pathway from genes to connectivity and ultimately, to function. This project will also integrate other brain dimensions by performing multidimensional network correlative analytics. In addition, this project will address the relationship between gene expression, cell types, and brain structures. The success of this project will be a new class of efficient, robust analytics methods that are flexible enough to be adapted for integrating, modeling, and mining current and future brain data.The results of this project will have an immediate and strong impact on multiple disciplines, including brain data analytics and computational neuroscience, biological image informatics, and big data analytics. A future long term goal is to uncover basic underlying differences between normal and impaired brain functions. The unified treatment of brain data analytics will be readily transformable into new courses for training next-generation computational biologists. The multidisciplinary nature of this project provides unique opportunities for integrating its components into existing curricula. Brain science has been shown to be a valuable resource for inspiring scientific interests in K-12 students. Components of the project will be integrated into an existing high-school student internship program, thereby inspiring future science students. Underrepresented students will be especially encouraged to participate throughout the project. The results of this project will be disseminated in the form of peer-reviewed publications, open-source software, tutorials, seminars, and workshops. All findings, publications, software, and data will be made publicly available at the project website: http://compbio.cs.odu.edu/CAREER/