TissueDistributionDBs: a repository of organism-specific tissue-distribution profiles

Tissue-distribution profiles are crucial for understanding the characteristics of cells and tissues in terms of their differential expression of genes. Most of the currently available resources for tissue-distribution profiles are either specialized for a few particular organisms, tissue types and disease stages or do not consider the "tissue ontology" levels for the calculation of the tissue-distribution profiles. Therefore, we have developed "TissueDistributionDBs", a repository of tissue-distribution profiles based on the expressed sequence tags (ESTs) data extracted from the UniGene database by employing "Tissue Ontology" available at BRENDA. To overcome the occurrence of the natural language variations in the EST's source tissue-type terms, we have generated a "tissue synonym library" and standardized these tissue-type terms by cross-referencing to the controlled vocabulary for tissue-type terms available at BRENDA "Tissue Ontology". Furthermore, we have provided a quantitative expression for genes among the tissue types at various anatomical levels by constructing "tissue slims". Concurrently, the expression among tissue types is used for tissue-distribution calculations. The resulting output profiles can be queried by the Sequence Retrieval System (SRS) and are currently available for 20 different model organisms. We benchmarked our database system against the Swissprot database using a set of 40 different tissue types. This database system is useful for the understanding of the tissue-specific expression patterns of genes, which have implications for the identification of possible new therapeutic drug targets, in gene discovery, and in the design and analysis of micro-arrays. TissueDistributionDBs can be accessed via the World Wide Web (www) at "http://genius.embnet.dkfz-heidelberg.de/menu/tissue_db/".