The purpose of sequencing the mouse genome was to further the career of Mus musculus as the biologist's favourite model of human disease. The task was completed in 2002, a year after the human genome. To exploit the new knowledge, a catalogue of mutant mice had to be created in the service of biomedical science. The outstanding questions were just how many genes needed to be individually mutated in mice, and how to set about it.

Some five years later, genetic technologies have developed so fast that the questions have virtually answered themselves. The community at large, in the form of the newly created International Mouse Knockout Consortium, has now declared that each and every one of the 20,000 or so genes in the mouse genome will be systematically targeted and mutated in embryonic stem cells. And all this is only the beginning.

The consortium is now taking requests from the community for genes to be targeted, with the gaps to be filled in later. Soon, if all goes according to plan, anyone will be able to order an off-the-shelf mutant mouse to test any biological hypotheses or develop any disease model.

The consortium formally launched itself earlier this month with the signing of a cooperation agreement between three funding agencies that together are committing several hundred thousand dollars to the cause over the next five years or so: the European Commission, the US National Institutes of Health and Genome Canada. The signing took place during a two-day meeting organized by the European Commission at a lakeside chateau in Genval, a village just outside Brussels, where delegates from around the world were able to discuss the implementation of the ambitious plan and to dream about the next steps.

The practicalities don't depend only on money. Databases are needed so that the consortium can efficiently share information and avoid duplication. The most important mutants need to be 'phenotyped', or characterized, to record the physiological effects caused by the lack of a gene. This means expanding and standardizing the activities of the 'mouse clinics' that have sprung up, mostly in Europe, to support previous research programmes. The question of how much phenotyping needs to be done during this phase, and on how many of the chosen mutants, still remains to be resolved.

Grandiose as these plans are, they are but one major step towards the vision of offering an even fuller service to biologists. For example, most of the embryonic stem-cell mutants currently available are 'null' knockouts — the targeted gene simply doesn't function. But, at a greater cost, it is now possible to make 'null-first conditional-ready' mutants. In these, the gene is knocked out by default but can be re-established and knocked out at will in particular tissues at particular times. This flexibility is much more valuable to researchers.

This technology cannot currently be applied to all genes, but it is developing fast. A fuller service would require that more extensive phenotyping be done on each of the mutants. Moreover, a further database is required to document the differences between mouse and human gene function, to ensure a deeper understanding of mouse models of human disease. The full service will be costly.

This vision represents the fulfilment of mouse genome sequencing. Support for that project needs to be followed through: the mouse has already led to excellent insights into many human diseases, and the continuation of this approach will deliver many more. Budgets have tightened, but funding agencies that stay the course can be assured of ample returns on their investment.