Collaborations between the public and private sectors can work well, as exemplified by the SNP Consortium, which recently made a dense map of human single-nucleotide polymorphisms (SNPs) available to the public. Allen Roses reviews how SNP-mapping technologies can identify patterns of susceptiblity genes, as well as efficacy and risk factors, which could contribute to reducing attrition in drug development. SNP technology is just one example of how new tools are changing drug discovery. Matthew Cooper reviews advances in optical biosensors, which exploit surface plasmon resonance to analyse biomolecular interactions. Developments have led to biosensor use in diverse areas of drug discovery. In the Perspectives section, Graham Richards focuses our attention on the screensaver project, which uses the power of grid computing to screen a virtual library of 3.5 billion molecules against cancer targets. This issue features the first in our new series of Case Histories, which will examine key milestones in the discovery and development of particularly important drugs. Capdeville and colleagues consider the case of Glivec, the first selective tyrosine-kinase inhibitor approved for the treatment of a cancer. In the field of nucleic-acid therapeutics, the ability to target selectively is crucial. Gewirtz and Opalinska review the different strategies for modulating gene expression, highlighting the clinical successes as well as the problems. In his review on insulin delivery, David Owens critically appraises the potential for alternative delivery routes for insulin that might reduce the need for injection. The pulmonary route of administration seems particularly promising, and could have important implications for the less-invasive delivery of peptides and proteins in general.