FGE-OM is an object model for capturing data from functional genomics experiments. Currently, the model covers microarrays and proteomics but will be extended in the future to include other technologies including: immunohistochemistry, metabolomics and protein arrays.

FGE-OM is divided into three namespaces:

• BioOM - covering technologies generic to any functional genomics experiment
• ArrayOM - technologies specific to microarrays
• ProteomicsOM - technologies specific to proteomics

The model is an extension of MAGE-OM, the basis for the standard format for microarrays. BioOM and ArrayOM contain classes derived from MAGE-OM. The ProteomicsOM namespace is derived from the existing draft standard for proteomics: PEDRo and a proposal from the University of Glasgow: Gla-PSI, Comparative and Functional Genomics, 2003. FGE-OM is expressed in UML (Unified Modeling Language).

A prototype database implementation has been developed in collaboration with the University of Glasgow. Tables covering BioOM and ArrayOM have been derived from the RAD schema at Penn. Proteomics data is stored in tables derived from the PEDRo schema and Gla-PSI.

A database implementation of FGE-OM has been created in Oracle known as RAPAD. Data is loaded using web pages adapted from the RAD Study Annotator, and 2-DE data can be viewed using a Java applet. Screenshots of data submission pages and the 2-DE viewer can be viewed. The proteomics component of the database will ultimately be incorporated into GUS. The current database schema is capable of storing data, protocols and analysis from microarrays and proteomics. In future versions of GUS, technology specific components will be stored separately in RAD and ProtDB.

Download schema (release Sep 2003) for Oracle version 9.