ABI Innovation: Next-Gen Clustering: Fast and Accurate Ways to Cluster Proteins

DNA测序技术降低了成本，提高了基因组测序的速度，这使得测序基因组的可用性呈指数级增长。为了充分利用这种情况，研究人员必须拥有易于使用但功能强大的软件工具。不幸的是，开源软件的供应商一直在努力跟上每天产生的海量生物数据的步伐。因此，许多研究人员正在使用工具来完成软件不打算执行的任务，导致结果不理想。需要合适的开源软件工具来处理大量的基因组数据，并获得科学突破所需的知识。该计划的智力价值在于，它将填补一个重要的空白，即需要一个软件程序，可以使用云计算快速准确地从测序的基因组中聚集数百万个蛋白质序列。从这些蛋白质簇中，系统基因组学方法可用于预测未表征蛋白质的功能。已经开发了几个版本的软件工具pClust，所获得的结果远远优于通过任何其他方式生成的结果。为了完成这项工作，我们将开发一个版本的pClust，以便在云中使用，并支持增量集群。此外，pClust将用于从所有可用的全基因组序列中对整个变形菌门（目前约1770种）的蛋白质进行聚类。这将是迄今为止对变形菌门最全面的研究，并将代表一个从未实现的规模。这个项目将在几个不同的领域产生更广泛的影响。当通过云计算和用户友好的图形用户界面可用时，pClust程序的并行配置将使全世界的科学家能够同时分析数千个基因组，快速准确地聚集基因组中的所有蛋白质。该软件工具的部署将对科学和医学的发现产生重大影响。在教育方面，将同时培养研究生和本科生。重点将放在招收和吸引女学生参与这个跨学科项目上，项目团队将参加每年面向初高中学生的拓展项目。本项目的主要研究成果将发表在同行评议的期刊和会议上。此外，作为该项目的一部分开发的软件工具将在谷歌Code上作为开源发布，聚类的结果，包括蛋白质簇本身、系统发育树、系统发育谱和系统发育谱之间的百分比一致性将在WSU电子工程和计算机科学学院的生物信息学和计算生物学网站上免费提供。

DNA sequencing technology has decreased the cost and increased the speed of genome sequencing, and this has led to an exponential growth in the availability of sequenced genomes. In order to take full advantage of the situation, researchers must have easy-to-use but powerful software tools. Unfortunately, it has been a struggle for providers of open-source software to keep pace with the immense volumes of biological data generated on a daily basis. As a result, many researchers are using tools for tasks the software was not meant to perform resulting in less than optimal results. Suitable open-source software tools are needed to process the quantities of genome data and obtain the knowledge needed for scientific breakthroughs. The intellectual merit of the proposed project is that it will fill one significant void, the need for a software program that can be used to cluster millions of protein sequences from sequenced genomes quickly and accurately using cloud computing. From these protein clusters, a phylogenomics approach can be used to predict the function of uncharacterized proteins. Several versions of the software tool pClust have already been developed, and results obtained are far superior to those generated by any other means. For this work a version of pClust will be developed for use in the cloud and with support for incremental clustering. In addition pClust will be used to cluster proteins from all available whole genome sequences for the entire bacterial phylum Proteobacteria (approximately 1770 species presently). This will be the most comprehensive study of the Proteobacteria phylum ever and will represent a scale that has never as yet been achieved.The broader impacts of this project will occur in several different areas. When made available via cloud computing and a user-friendly graphical user interface, the parallel configuration of the pClust program will allow scientists worldwide to analyze thousands of genomes at once, quickly and accurately clustering all the proteins within the genomes. Deployment of this software tool will have a significant impact on discoveries in science and medicine. In terms of education, both graduate and undergraduate students will be trained. Emphasis will be placed on recruiting and engaging female students into this interdisciplinary project, and the project team will participate in annual outreach programs to middle and high school students. Key research findings originating from this project will be published in peer-reviewed journals and conferences. Also, the software tools developed as part of this project will be published as open source at Google Code, and the results of the clustering, including the protein clusters themselves, phylogenetic trees, phylogenetic profiles, and percentage agreement among the phylogenetic profiles will be made freely available on the WSU School of Electrical Engineering and Computer Science's Bioinformatics and Computational Biology Web site.