Cloud Based Resource for Data Hosting, Visualization and Analysis Using UCSC Canc

描述（申请人提供）：癌症基因组资源正以前所未有的速度增长。然而，对癌症基因组的全面分析仍然是一项艰巨的挑战。这部分是由于目前的技术在可视化、整合和分析癌症基因组数据方面存在困难。我们建议开发一个基于云的平台，使研究人员能够托管、可视化和分析他们自己的数据。该平台由一组癌症分析虚拟机（cavm）组成。每个CAVM的主要组件是一个数据服务器，其功能是存储用户数据并将其提供给应用程序，例如UCSC癌症基因组学浏览器，以提供数据可视化。第二个组件是经过修改的Galaxy工作流系统，用于提供数据分析功能。UCSC用于下一代测序数据分析和途径推断的分析工具套件将与该系统一起预先打包。这两个组件将高度集成，以实现数据可视化和分析的紧密耦合周期。数据服务器组件将是模块化的，这样它就可以独立地向除Cancer Browser和Galaxy之外的应用程序提供数据。我们将提供虚拟机映像，这些映像可以很容易地在商业云（如Amazon）中启动，也可以安装在用户自己的机构中。CAVM还可以作为用户与外部大型数据库集成的一种方式。我们将提供一个UCSC CAVM，其他CAVM实例可以连接到该CAVM，以提供来自UCSC癌症基因组数据库的授权数据访问。该系统允许动态形成由多个来源的数据片组成的新数据集。这种将数据组合成更大样本的能力，将提供统计学上的力量，使发现成为可能，否则是不可能的。我们的目标是消除或显著减少系统配置和软件安装的开销。我们的工具将为用户提供访问基于云的集群计算环境的能力，这将使那些无法访问计算服务器的研究人员能够访问复杂的、计算密集型的分析。我们开发的软件平台可以被个人实验室生物学家使用，也可以被大型项目用于向个人用户或其他项目提供数据。这种设计有可能形成可通过相同软件接口访问的扩展联邦数据库。

DESCRIPTION (provided by applicant): Cancer genomics resources are growing at an unprecedented pace. However, a comprehensive analysis of the cancer genome still remains a daunting challenge. This is in part due to the difficulties in visualizing, integrating, and analyzng cancer genomics data with current technologies. We propose to develop a cloud-based platform to empower researchers with the ability to host, visualize and analyze their own data. The platform is composed of a set of Cancer Analytics Virtual Machines (CAVMs). The main component of each CAVM is a data server which functions to store and serve user data to applications, such as the UCSC Cancer Genomics Browser, to provide data visualization. The second component is a modified Galaxy workflow system to provide data analysis capability. UCSC's suite of analysis tools for nextgen sequencing data analysis and pathway inference will be prepackaged with the system. The two components will be highly integrated to allow tightly coupled cycles of data visualization and analysis. The data server component will be modular such that it can provide data independently to applications besides the Cancer Browser and Galaxy. We will deliver virtual machine images that can be easily initiated in a commercial cloud such as Amazon, or can be installed within a user's own institution. The CAVM also functions as a way for users to Integrate with external large-scale databases. We will deliver a UCSC CAVM that other CAVM instances can connect to, to provide authorized data access from the UCSC cancer genomics data repository. The system allows the dynamic formation of new datasets composed of data slices from multiple sources. This ability to combine data into larger samples will provide the statistical power to allow discoveries that would otherwise not be possible. We aim to eliminate, or significantly reduce, the overhead of system configuration and software installation. Our tools will provide users the capability to access a cloud-based cluster computing environment, which will make sophisticated, computationally intensive analyses accessible to researchers who might not, have access to compute servers. The software platform we develop can be used by individual bench biologists, and also by large projects to serve data to individual users or to other projects. This design has the potential to form an expansive federated database accessible through the same software interface. RELEVANCE: Currently, clinicians and bench biologists typically depend on external collaborators for data analysis. The proposed system will provide these scientists with data analysis and visualization methods that are both powerful and easy to use. This will accelerate research in the understanding and treatment of cancer, the second-leading cause of death in the U.S.