washington

Leaders of the publicly funded Human Genome Project have rejected a proposed memorandum of understanding between one of the project's participants, the US Department of Energy (DoE), and Celera, the corporation established by geneticist Craig Venter in a rival bid to sequence the genome.

There had been concern that clauses in the proposed agreement would have been in conflict with the desire for open access to sequence data. But the leaders of the government project still hope that the DoE can reach an agreement with Celera before the corporation starts its sequencing work this spring (see Nature 393, 101; 1998).

The National Human Genome Research Institute (NHGRI) at the US National Institutes of Health (NIH) and the UK Wellcome Trust — the two largest participants in the international genome project — last month scotched the draft agreement under which Celera and the DoE would have exchanged sequence information on the three chromosomes of the genome that are being sequenced by the DoE.

Francis Collins, director of the NHGRI, says his institute and the trust were “uncomfortable” with the draft because it would have required DoE scientists to work with data from Celera that was not publicly available, and it involved the government agency in providing a service to a single, profit-making corporation. The draft would have been “in violation of the international consensus to have sequence data released immediately” into the public domain.

Collins and Michael Morgan, a senior official responsible for the genome project at the Wellcome Trust, asked Ari Patrinos, head of biological and environmental research at DoE, to come up with a new agreement.

Patrinos now concedes that the draft would have “given some scientists an early peek” at genome data, and that this was “inappropriate and incorrect”. Patrinos expects that a new agreement will be negotiated.

Committees of the US Congress have pressed the NIH and DoE to collaborate with Venter's effort, but differing approaches to the release of data will make this easier said than done. The government project requires all sequence material to be published immediately. Celera wants to keep its data secret for three months — scouring it for information that can be patented — before sharing it with the outside world.

Collins says he is unsure how the agreement can be made to work, but says that “no data will be exchanged that is not publicly available”.

Paul Gilman, head of policy at Celera, says the corporation will be receptive to any new proposal from the public agencies, but that it “hadn't heard a word from them” since the draft was withdrawn in December.

Collins strongly denied a report in Timemagazine that “sour grapes” at the NIH was behind the rejection of the draft. Venter has a history of friction with the NIH, which he left in 1992 after an explosive public argument over gene patents.