IOBIO: Web-based, interactive tools for real-time analysis in genomic big data

基因组分析有可能彻底改变遗传性疾病、传染病和癌症的方式 得到了诊断和治疗。然而，现有的基因组分析工具已经针对处理进行了优化 端到端的全基因组或外显子组数据集，需要生物信息学培训，昂贵的计算机 硬件和数据存储，使得许多没有 获取这些资源的途径。这些工具需要几个小时或几天来完成分析，并产生大型、静态的 需要大量专业知识才能解释的输出文件。这种详尽的“自上而下”的方法并不 为个别研究人员提供快速检查和排除数据集故障的方法，或测试 在法官席上或在临床上提出的假说。因此，现有的基因组分析工具不足以 接触到最终用户，例如研究临床医生，他们最需要他们来影响基因组方面的重大进展 医学，但由于他们的临床职责，他们有最少的时间和最零碎的时间 研究。我们正在开发iobio(http://iobio.io)，)，这是一种新型的基因组分析系统，将使 生物医学专业人员无需计算资源即可访问，并交互分析生物医学大数据 仅使用一台笔记本电脑，就可以获得基因组级别的数据。而不是分析完整的基因组数据集结束 最后，每个iobio应用程序执行集中的基因组分析(例如，在基因区域中)，并返回 结果以秒为单位。使用复杂而直观的Web界面直观地显示结果，允许 科学家使用Web服务器版本快速处理他们的数据，这些工具与在END中使用的功能强大的UNIX工具相同 端到端基因组分析，但不需要计算硬件和工具安装，可视化其 结果，展开或细化并立即重复以定制其分析策略。Iobio工具包 (http://iobio.io)目前包括四个功能齐全的Web应用程序，已经为复杂的继承变体提供了便利 优先排序和元基因组分析。我们不断增长的用户群已经有数千人，其中许多人 返回那些将我们的应用程序整合到他们的分析例程中的“客户”。在这里，我们建议大大地 扩展我们现有的工具箱，以支持癌症基因组研究。癌症基因组是高度 根据患者的不同而变化，需要可定制的分析，非常适合我们的交互式iobio Web工具。意识到仅靠我们的团队将不能为每个任务开发和维护工具 基因组学研究的子领域，我们将构建广泛的软件库来支持iobio应用程序的开发 由第三方开发人员小组提供。我们还将开发灵活的选项，以便高效地操作我们的工具 在本地服务器硬件上或在计算云环境中安全地运行。Iobio将成长为一个富有的 充满活力的分析生态系统，将增强各级生物信息学专业水平的生物医学研究人员的能力， 精通计算的研究人员和板凳科学家，执行难以进行的直观数据分析 用现有的基因组分析工具完成。

Genomic analyses have the potential to revolutionize the way inherited disease, infectious disease, and cancer are diagnosed and treated. However, existing genomic analysis tools have been optimized for processing whole genome or exome datasets from end to end, requiring bioinformatics training, expensive computer hardware and data storage, rendering them unavailable to many biomedical researchers who don't have access to such resources. These tools take hours or days to complete analysis, and produce large, static output files that require considerable expertise to interpret. This exhaustive “top-down” approach does not provide individual researchers with the means to quickly examine and troubleshoot their datasets, or test hypotheses formulated at the bench or in the clinic. Thus, existing genomic analysis tools do not adequately reach the end users, e.g. research clinicians who need them most to affect major advances in genomic medicine, but because of their clinical duties have the least amount and most fragmented time for their research. We are developing iobio (http://iobio.io), a novel genomic analysis system that will enable biomedical professionals without computational resources to access, and interactively analyze biomedical big data at the genome scale, using only a laptop computer. Instead of analyzing complete genomic datasets end to end, each iobio app performs focused genomic analyses (e.g. in the region of a gene) and returns the results in seconds. Results are displayed visually using a sophisticated and intuitive web interface, allowing scientists to quickly process their data using web server versions of the same powerful UNIX tools used in end- to-end genomic analyses but without the need for computing hardware and tool installation, visualize their results, expand or refine and immediately repeat to customize their analysis strategy. The iobio toolkit (http://iobio.io) currently includes four full-featured web apps, already facilitating sophisticated inherited variant prioritization and metagenomic analyses. Our growing user base is already in the thousands, many of them returning “customers” who have incorporated our apps into their analysis routine. Here, we propose to vastly expand our existing tool chest for supporting cancer genomic investigation. Cancer genomes are highly variable from patient to patient, requiring customizable analyses, tasks ideally suited for our interactive iobio web tools. Realizing that our team alone will not be able to develop and maintain tools for every task in every subdomain of genomics research, we will build extensive software libraries to support iobio app development by third-party developer groups. We will also develop flexible options for operating our tools efficiently and securely, on local server hardware or in computational cloud environments. iobio will grow into a rich and vibrant analysis ecosystem that will empower biomedical researchers at all levels of bioinformatics expertise, computationally skilled researchers and bench scientists, to carry out intuitive data analyses that are difficult to accomplish with existing genomic analysis tools.