CD-HIT: A Fast Program to Cluster and Compare Large Sets of Biological Sequences

描述(由申请人提供)：项目摘要/摘要CD-HIT是一种用于对大量蛋白质或核苷酸序列进行分类和比较的计算机程序。它有助于显著减少各种序列分析任务中的计算和手动工作，并有助于理解数据结构和纠正数据集中的偏差。CD-Hit比其他方法快2到3个数量级。它可以处理极大的数据库，并在各个领域得到了广泛的应用。根据用户的反馈和越来越多的出版物引用CD-Hit，CD-Hit正变得越来越受欢迎。CD-HIT现在有数千名用户，经常用于许多流行的数据库，如UniProt和PDB。由于高通量基因组测序计划和最近的环境元基因组计划，公共序列数据库的爆炸性增长使研究人员面临着严重的挑战和问题。常规分析，从搜索数据库到建立多个比对，计算成本越来越高，也越来越复杂。一种高效的聚类方法对于解决许多挑战和帮助研究人员克服这些问题至关重要。目前，就速度和处理非常大的数据集的能力而言，没有其他可用的程序可以取代CD-HIT。因此，CD-Hit将在未来发挥更重要的作用。这项提议的目标是进一步改进和发展CD-HIT方案和相关应用程序，以便更好地服务于日益增多的用户群体，并解决CD-HIT用户提出的问题。该算法将进行改进，以达到更好的性能，并克服现有的局限性。在保持超高速的同时，将努力实现更准确的聚类结果。将实施新的功能，以满足各种集群和比较需求。将出现更多增强的维护和更好的软件工程技术，以提供定期的程序发布和更新、更好的可移植性、更短的故障排除周期和更丰富的文档。根据大学的政策，CD-Hit将继续是一个开源程序包。此外，还将设置一个网络服务器，以便公众更容易访问CD-HIT的应用程序。该服务器将提供进一步的分析和可视化工具、界面和其他生物信息学资源的链接。预先计算的流行数据集将向公众提供，以消除各个实验室重复相同工作的需要。项目叙事CD-HIT是一个快速的计算机程序，用于对公共卫生相关研究中的数千名研究人员使用的生物序列进行分类和比较。它直接帮助研究人员大大减少了序列分析的工作量，并纠正了公共数据库中的偏差。随着公共序列数据库的爆炸性增长，CD-HIT的持续发展将更好地服务于在序列分析方面面临更多挑战的研究人员。

DESCRIPTION (provided by applicant): Project Summary/Abstract CD-HIT is a computer program for clustering and comparing large sets of protein or nucleotide sequences. It helps to significantly reduce the computational and manual efforts in various sequence analysis tasks and aids in understanding the data structure and correct the bias within a dataset. CD-HIT is 2 to 3 orders of magnitude faster than other methods. It can handle extremely large databases and has been used extensively in various fields. CD-HIT is becoming increasingly popular based on users' feedback and the growing number of publications that cited CD-HIT. CD-HIT has thousands of users now and is routinely used in many popular databases, such as UniProt and PDB. Researchers are now facing serious challenges and problems from the explosive growth of public sequence databases as a result of high-throughput genome sequencing projects and the very recent environmental metagenomic projects. The routine analysis, from searching a database to building a multiple alignment, is getting more computational expensive and complicated. An efficient clustering method is crucial to address many of the challenges and help researchers to overcome the problems. Currently, no other available program can replace CD-HIT in terms of speed and the ability to handle very large datasets. Therefore, CD-HIT will be playing a more important role in the future. The goal of this proposal is the further improvement and development of the CD-HIT program and related applications to better serve the increasing user community and to address the issues raised by users of CD-HIT. The algorithm will be improved to achieve better performance and overcome the existing limitations. Efforts will be spent towards more accurate clustering results while still maintaining the ultrahigh speed. New functions will be implemented to meet various clustering and comparing needs. More enhanced maintenance and better software engineering techniques will take place to provide regular program releases and updates, better portability, shorter trouble shooting cycles, and richer documentation. Subject to University policies, CD-HIT will be continually an open source package. In addition, a web server will be set up for easier public access to CD-HIT's applications. The server will provide further analysis and visualization tools, interface and links to other bioinformatics resources. Pre-calculated popular datasets will be made available to the public to eliminate the need for individual labs to repeat the same work. Project Narrative CD-HIT is a fast computer program for clustering and comparing biological sequences used by thousands of researchers in public health related studies. It directly helps researchers to significantly reduce the efforts in sequence analysis and to correct the bias within public databases. Continued development of CD-HIT will better serve researchers who are facing more challenges in sequence analysis by the explosive growth of public sequence databases.