OAT - The Ontology Annotation Tree browser tool The microarray technique has gained enormous popularity both among the major pharmaceutical companies and academic institutions. Even though all the benefits of the technique, we have to face the problems of that the amount of resulting information of microarray analysis is of such magni- tude that it is difficult to get an overview of the data. The need of condensing the information of an analysis is therefore evident. In order to address this challenge we have developed the Ontology Annotation Tree browser tool (OAT-tool). The OAT-tool utilizes the two ontologies of Medical Subject Headings (MeSH) and Gene Ontology (GO) to represent the information. Motivation for OAT Link to MeSH description Number of annotations below Number of probe set below Number of probe set annotated with this term Anders Bresell 1, Bo Servenius 2 1.Department of Computer and Information Science, Linköpings Universitet, Linköping, Sweden. 2.Department of Molecular Sciences, AstraZeneca R&D Lund, Lund Sweden. URL: 1. Submit probe sets 2. Browse the Ontology 3. Make a report Checkbox for terms which are to be included in the report Significance of the Annotation The goal of the Gene Ontology TM Consort- ium is to produce a dynamic controlled vocabulary that can be applied to all organ- MeSH is the National Library of Medicine's controlled vo- cabulary thesaurus. Thesauri are carefully constructed sets of terms often connected by broader-than, narrower-than, treesetup.cgi OATdb treebrowser.cgi QSdb report.cgi annotationReport.cgi geneReport.cgi (a) (b) (c) (d) (f) (g) (h) (i) (j) (k) (e) At the home page of OAT (a) a link to the query form is found. (b) The probe sets and the tree option is sent to the tree setup script. The relevant information of the query is (c) extracted from OATdb and (d) stored in QSdb. The tree browser script (e) reads data from QSdb and (f) visualises it as a web page. For each modification of the tree visualisation (g) the tree browser script is reloaded with the new data. By marking terms and clicking on the submit button (h) the report information is sent to (i) a redirection script. The report scripts, one each for the sorting of data in terms of genes or term strings, (j) reads data from QSdb and (k) generates a report web page. Technical Components The OAT-tool runs from a web server. It is written in Perl and CGI. All data are stored in the OAT database (OATdb) and for each query a Query Specific database (QSdb) is generated. The information flow of OAT’s three main scripts is outlined below to the right. The gene information is collected from Affymeterix TM -files. Ontology information is collected from MeSH and GO. Annotations is collected from EMBL/MEDLINE and Gene Ontology Annotation Campaign (GOAC). The source of information Link to MEDLINE Link to Affymetrix DB Link to MeSH DB The information flow of OAT and related links. These links show the relationships between related terms and provide a hierarchical structure that permits searching at various levels of specificity from narrower to broader. There are more than 19,000 terms in MeSH isms. The terms are structured in three ontologies: Molecular Function, Biological Process and Cellular Componet. For each query of probe sets we extract a subset of the ontology and assign the annotated genes to the different terms in the ontology. The user browse the ontology in a hierarchical way from the top-level terms and down to the more detailed ones. When a satisfying level of detail is reached the user have the ability of summarizing the information in a report which could be used for further studies Work flow of the OAT-tool A more detailed description of the work is given in Interpretion of microarray expression data using ontology browsing, a master thesis report (LiTH-IDA-Ex-02/75) from Linköpings Universitet by Anders Bresell.