Genquire under the hood: live database-driven graphical objects David Block National Research Council Plant Biotechnology Institute Saskatoon, Saskatchewan In order to allow real-time annotation of biological sequences, it was necessary to create an object framework with all the functionality of BioPerl SeqFeatures, but stored in a relational database. Starting from example code from 'Object Oriented Perl' by Damian Conway(1), such a framework was implemented, first without persistence, then with a MySQL backend. The GFF framework is the basis for the storage of features, with a flexible tag-value scheme added to provide extensibility. The actual text of source sequences is stored in the database, with optimized loading and caching for better client performance. The presence of a Tiling_Path table allows the loading of any arbitrary length of sequence, in the reference frame of any of the contigs that make up the tiling path. This functionality is used in the genome browser window of Genquire. Hand annotation requires mutable sequence features, but it was thought important to preserve original data from annotation pipelines. Feature::mod was developed to allow such modification. Annotation was conceived of as a process of defining gene structure, so a Gene class, with associated Feature::Annotation exon objects, was defined. These classes have built-in logging of every change, which theoretically allows for Undo functionality. Gene Ontology annotation is seen as the future of biological integration, and GO annotations are stored simply and completely by these Feature Objects. Extension of these classes back away from database integration was trivially implemented as Feature::inMemory, which will be demonstrated. Current work on adding transaction support, Corba client/server integration, and DAS will be presented. The trend in this work is to separate the biological logic from the database implementation, and a strategy to this end will be presented. 1. Conway, Damian. Object-Oriented Perl. Manning, 1999. -- David Block dblock@gene.pbi.nrc.ca http://bioinfo.pbi.nrc.ca/dblock/wiki NRC Plant Biotechnology Institute Saskatoon, SK, Canada