The international Human Proteome Organisation (HUPO) was launched in 2001 to much scepticism. After all, the proteome — the complete set of proteins expressed by a cell's genome, and modified following expression — is a moving target. Its constituents change continuously according to the conditions to which the cell is exposed. A ‘human proteome project’, in contrast to the Human Genome Project, would be a fuzzy, infinite endeavour. What role could there be for an international organization?

Proteomics was in any case enjoying a mixed reputation. The work of a small number of excellent labs was being diluted by vast data dumps of dubious value. The new techniques developed to identify proteins on a large scale were being snapped up by inexperienced users who pumped out data that often proved hard to reproduce. Even in the best hands, the various techniques had very different outputs, making comparison of results between labs difficult. And it was clear to many that the simple cataloguing of long lists of proteins was not going to carry biology very far forward. The boom was threatened by a bust.

No longer. As last week's fourth annual HUPO meeting in Munich demonstrated, the proteomics community is finally putting its house in order. And HUPO itself is playing an indispensable role. Formally established with a permanent secretariat in Montreal and supported in part by the Canadian government, it has set up committees whose members — the scientific heavyweights of proteomics — select initiatives to come under HUPO's wing. Such projects will now be required to maintain HUPO's standards, and will be supported by HUPO's training and education programmes.

Seven initiatives have already been chosen, including proteome projects in the brain, liver and plasma. Crucially, a major focus will be HUPO's Proteomics Standards Initiative, which is developing standards for data generation and presentation, including standardized formats for databases of the full range of proteomics measurements.

Without the umbrella of HUPO, researchers would have been more inclined to use their rivals' toothbrushes than their protocols.

For example, protein–protein interaction data have been generated using both the ‘yeast two-hybrid’ system and mass spectroscopy, and held in different databases. The initiative has agreed on a common database format that sends users to information in both databases when asked which other proteins a specified protein interacts with.

For mass spectroscopy data, the initiative is developing a list of information that a researcher must provide to accompany a claim that a protein has been identified, including the names of fragments and their individual masses. HUPO may eventually decide that the spectra themselves must be provided.

Two related activities represent important new resources. One is a proteomics database called PRIDE (, developed jointly by the European Bioinformatics Institute at Hinxton near Cambridge, UK, and Ghent University in Belgium, which will use the HUPO standardized formats.

The second reflects HUPO's commitment to a broader definition of proteomics in the context of systems biology — using proteomic techniques to follow particular proteins in a biological pathway. Five major protein-interaction databases have agreed to share curating efforts and provide information in a HUPO standardized format, in a consortium called the International Molecular Exchange (IMEX).

Without the umbrella of HUPO, hopes for standardization in proteomics would have been bleak, with researchers being more inclined to use their rivals' toothbrushes than their protocols. HUPO is involving the entire international community in its discussions to ensure consensus, and has already brokered a surprising number of agreements, with journals ready to assist in enforcing standards.

Even as HUPO brings order to chaos, researchers are proving the value that proteomics is bringing to biology, not just in identifying biomarkers useful in medicine, but also in understanding how relatively simple genomic information is transformed into the wonder that is the functioning cell. At the end of the day, proteins, not genes, are the business end of biology.