Conceptual comparison of COLOMBOS with similar initiatives.

1Compendium: a data matrix (genes in rows, microarrays in columns) combining expression measurements from different experiments (an experiment being a set of microarrays submitted to the public DBs as such, implying that they were performed by the same lab and on the same technological platform). Single- vs. cross-platform: combining data from the same technological platform is relatively easy as the same preprocessing methodology can be employed; COLOMBOS is unique in combining data from different platforms using a specialized homogenization pipeline. Meta-analysis: expression data are not combined directly but experiments are analyzed separately where after the results are compared. 2The biological conditions measured on a microarray are described with a set of formal terms which are organized into a higher level ontology. Such an ontology facilitates querying for related experiments or conditions. 3Single versus multiple queries: query results can be retained in the COLOMBOS user workspace where they can be organized and structured, into larger ‘analysis projects’. This allows for integrative across-query analysis where relations between single query results can be explored, e.g. by combining or differentiating single query results.