We live in the age of information overload. This statement applies to all walks of life, and in genetics we have come to realize that it is not the data acquisition but the analysis that is the next big problem to be solved. Moreover, it is often necessary to analyse different types of data together to get the complete picture. The highlight on page 330 discusses one such example, in which genetic screens of gene expression in mouse, maize and human were performed using a combination of expression data and genome-wide markers to map loci that control gene expression.

But data are also gathered simultaneously by different labs and the way each lab or centre decides to manage their data is often unique. The quantity of data (and the speed with which new data are acquired) means that keeping the various databases up to date can be problematic even within one centre. So, is the call for database integration just an unrealistic dream? Lincoln Stein discusses this issue on page 337. He uses familiar examples of databases such as Ensembl, WormBase and FlyBase to illustrate the problems that prevent database integration and proposes a solution for how these problems might be circumvented.

A topic that is indirectly related to the functional problems of data integration is that of private versus public databases. The highlight on page 333 discusses two recent papers that assess the quality and usefulness of the existing SNP databases. Although both groups conclude that the quality of SNP databases is high, and they both call for further SNP genotyping, especially from populations of non-European descent, some discrepancies between their results might be attributed to the fact that only one group had access to both private and public SNP collections.