Gene expression data mining guided by genomic background knowledge

Microarray data represents valuable information resources, nevertheless the knowledge is hidden inside the data and it is not easy to mine. Background knowledge is also stored in various formats and it is challenging to automatically infer the biological meaning from existing repositories. This paper deals with a new gene-expression knowledgefusion system that combines molecular biology data from various sources — the experiment in hand, gene expression data from similar experiments stored in array expression databases, additional knowledge on the most significant genes and their products from specialised services (e.g., pathway databases), and automatically derived results provided by relevant scientific literature. The design of the proposed system is rather complex. We take advantage of recent semantic web technologies to integrate the various modules of the system. Some of the components described in the paper have already taken part in the end-user applications, others still wait for their implementation in the form of software tools.