Home : Meetings : MGED1
An international meeting on Microarray Gene Expression Databases - MGED1
November 14-15, 1999
European Bioinformatics Institute,
Wellcome Trust Genome Campus, UK
The goal of the meeting was to create a framework for developing standards for storing and communicating microarray based gene expression data. The particular discussion topics of the meeting were defining the minimal information about microarray experiments that laboratories should be encouraged to provide to ensure reproducibility and verifiability of results, identifying where controlled vocabularies for annotating the samples and experiments are possible, and defining the data communication standards. Another important issue addressed by the meeting was developing standards for cross-platform data comparison and normalisation.
The meeting accepted a list of general recommendations, as well as established five working groups:
- Experiment description and data representation standards
- Microarray data XML exchange format
- Ontologies for sample description
- Normalisation, quality control and cross-platform comparison
- Future user group: queries, query language, data mining
The participants also agreed to meet again in 6 months time. The coordinators of the working groups jointly with representatives of some of the microarray groups involved in database development will be steering the overall initiative till next meeting.
The meeting also discussed ongoing microarray gene expression database projects (see agenda). Several of the largest microarray laboratories in academia: Stanford University, Whitehead Institute, the University of Pennsylvania, and National Cancer Institute are developing their own internal databases. A lot can be learned from internal database projects in the largest commercial microarray companies, like Affymetrix and Incyte, and commercial database projects, like the one by GeneLogic. Harvard University has created a database compiling and representing data in a consistent format from several public sources. There are at least three ongoing projects for creating public gene expression data repositories, namely, GeneX in National Center for Genome Resources (NCGR), Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI), both in the US, and ArrayExpress at the EBI. A similar project is also under consideration in DDBJ in Japan.
This diversity of gene expression databases may promote development, provided there is complete exchange of data. Discussions are ongoing between the various parties to ensure common standards to allow such data exchange and integration, and the meeting at the EBI created optimism that such standards will be agreed.