Automated and Portable Sequence Analysis Using Virtual Machines and Cloud Computing

最近，由于第二代测序技术的可负担性，测序费用大幅下降，导致越来越多的小基因组测序设施的建立。尽管序列产生的速度增加了，但获得支持高质量序列处理和分析的计算资源的机会并没有相应增加。特别是，一些新进入该领域的研究人员现在正在获得下一代测序平台，他们可能没有做好充分准备，无法充分利用他们自己的高通量测序设备。这项由W.Florian Fricke博士和Owen White博士(医学大学医学院基因组科学研究所)开展的项目正在为没有大量生物信息学资源的研究人员增加最先进的序列分析软件的可及性。该项目正在使用虚拟机器(VM)开发一个可移植和独立的软件包，其中包括现成的、开放源码的基因组分析工具。VM设计将提供两个主要优势，允许用户1)绕过复杂的软件安装和2)避免本地计算网络的性能瓶颈。VM包将在与所有计算机操作系统兼容的单个可执行文件中包括完全可操作的生物信息学管道，使进一步的软件安装变得不必要。大型序列数据的处理可以外包给称为计算云的大型分布式计算网络。VM包提供的分析协议将复制和扩展已建立的生物信息学协议，并包括用于全基因组和元基因组注释和比较分析的工具，包括序列组装、基因预测、功能注释、代谢途径重建和系统发育分类。这个开放源码和支持云的VM包的可用性将增加微生物基因组测序对广泛用户社区的可用性。VM包将作为正在进行的工作提供，至少有两个试用版和四个生产版。它将作为一个开源软件工具通过项目网页下载(http://clovr.igs.umaryland.edu/).The VM包正在通过出版物、会议演示文稿、项目网站和在线博客广泛宣传和记录。此外，还通过万维网举办了一次在线研讨会(网络研讨会)，讲授微生物基因组分析的基础知识，使用与VM程序包一起分发的一组测试序列数据。基因组分析可以为微生物研究的许多领域提供重大好处。下一代测序技术的发布促进了一种负担得起的非集中化微生物序列分析的新模式，使整个科学界受益。这个便携的、开源的微生物序列分析包为这个模型的成功做出了贡献。

-----------------------Less than 15 years after the first complete sequencing of a bacterial genome, sequence analysis has now become an integral part of nearly all research areas in biology. Recently, sequencing expenses have dropped sharply due to the affordability of second generation sequencing technology leading to the establishment of an increasing number of small genome sequencing facilities. Despite the increased rate of sequence generation, there has not been a commensurate increase in access to computational resources to support high-quality sequence processing and analysis. In particular, some of the investigators new to the field who are now obtaining next generation sequencing platforms could be insufficiently prepared to take full advantage of their own high-throughput sequencing devices. This project by Drs. W. Florian Fricke and Owen White (Institute for Genome Sciences, Univ MD School of Medicine) is increasing the accessibility of state-of-the-art sequence analysis software to researchers without extensive bioinformatics resources. This project is developing a portable and stand-alone software package, using Virtual Machines (VM), that is incorporating readily available, open source tools for genome analysis. The VM design will provide two main advantages, allowing users to 1) circumvent complex software installations and 2) avoid performance bottlenecks of local computing networks. The VM package will include fully operational bioinformatics pipelines within a single executable file that is compatible with all computer operating systems and makes further software installations unnecessary. The processing of large sequence data can be outsourced to large distributed computing networks called compute clouds. The analysis protocols provided on the VM package will replicate and extend established bioinformatics protocols and include tools for whole genome and metagenome annotation and comparative analysis, including sequence assembly, gene prediction, functional annotation, metabolic pathway reconstruction, and phylogenetic classification. The availability of this open source and cloud-enabled VM package will increase the usability of microbial genome sequencing to a broad user community. The VM package will be made available as a work in progress with at least two trial and four production releases. It will be available for download as an open source software tool through the project webpage (http://clovr.igs.umaryland.edu/).The VM package is being extensively advertised and documented through publications, conference presentations, the project website and an online blog. In addition, an online seminar (webinar) is being offered through the World Wide Web to teach the basics of microbial genome analysis using a test set of sequence data distributed together with the VM package. Genome analysis can provide significant benefits to many areas of microbial research. The release of next-generation sequencing technologies promotes a new model of affordable, de-centralized microbial sequence analysis with benefits for the entire scientific community. This portable, open source, microbial sequence analysis package is contributing to the success of this model.