ABSTRACT
For many years, molecular biology has addressed functional questions by studying individual genes, either independently or a few at a time. In spite of the intrinsic reductionism of this approach, an important part of our knowledge on functional properties and biological roles of genes and gene products was obtained in this way. However, in only a few years, the advent of high-throughput technologies has drastically changed the scenario. DNA microarrays constitute probably the paradigm among these technologies (Rhodes and Chinnaiyan, 2005) which are characterized for producing large amounts of data, whose analysis and interpretation is far from being trivial. Thus, the possibility of obtaining experimental measurements of thousands of genes in a fast and relatively cheap way has opened up new possibilities in querying living systems at the genome level that are beyond the old paradigm ‘one-gene-one-postdoc’. Relevant biological questions regarding genes, gene product interactions or biological processes played by networks of components, etc., can now for the first time be realistically addressed. Nevertheless, genomic technologies are at the same time generating new challenges for data analysis and are demanding a drastic change in the habits of data management and in the biological interpretation of the results. Dealing with this overabundance of data must be approached cautiously because of the high occurrence of spurious associations if the proper methodologies are not used (Ge et al., 2003).
