Cheung Kei-Hoi, Hager Janet, Pan Deyun, Srivastava Ranjana, Mane Shrikant, Li Yuli, Miller Perry, Williams Kenneth R
Center for Medical Informatics, Department of Anesthesiology, Yale University, New Haven, CT, USA.
Nucleic Acids Res. 2004 Jul 1;32(Web Server issue):W441-4. doi: 10.1093/nar/gkh397.
We have developed a universal web server application (KARMA) that allows comparison and annotation of user-defined pairs of microarray platforms based on diverse types of genome annotation data (across different species) collected from multiple sources. The application is an effective tool for diverse microarray platforms, including arrays that are provided by (i) the Keck Microarray Resource at Yale, (ii) commercially available Affymetrix GeneChips and spotted arrays and (iii) custom arrays made by individual academics. The tool provides a web interface that allows users to input pairs of test files that represent diverse array platforms for either single or multiple species. The program dynamically identifies analogous DNA fragments spotted or synthesized on multiple microarray platforms based on the following types of information: (i) NCBI-Unigene identifiers, if the platforms being compared are within the same species or (ii) NCBI-Homologene data, if they are cross-species. The single-species comparison is implemented based on set operations: intersection, union and difference. Other forms of retrievable annotation data, including LocusLink, SwissProt and Gene Ontology (GO), are collected from multiple remote sites and stored in an integrated fashion using an Oracle database. The KARMA database, which is updated periodically, is available on line at the following URL: http://ymd.med.yale.edu/karma/cgi-bin/karma.pl.
我们开发了一个通用的网络服务器应用程序(KARMA),它能够基于从多个来源收集的不同类型的基因组注释数据(跨不同物种),对用户定义的微阵列平台对进行比较和注释。该应用程序是处理各种微阵列平台的有效工具,包括由以下机构提供的阵列:(i)耶鲁大学的凯克微阵列资源中心;(ii)市售的Affymetrix基因芯片和点阵式微阵列;以及(iii)个别学者制作的定制微阵列。该工具提供了一个网络界面,允许用户输入代表单个或多个物种的不同阵列平台的测试文件对。该程序根据以下类型的信息动态识别在多个微阵列平台上点样或合成的类似DNA片段:(i)如果所比较的平台属于同一物种,则为NCBI - Unigene标识符;或者(ii)如果是跨物种的,则为NCBI - Homologene数据。单物种比较是基于集合运算实现的:交集、并集和差集。其他可检索的注释数据形式,包括LocusLink、SwissProt和基因本体论(GO),是从多个远程站点收集的,并使用Oracle数据库以集成方式存储。定期更新的KARMA数据库可通过以下网址在线获取:http://ymd.med.yale.edu/karma/cgi-bin/karma.pl。