London

Secret talks between rival teams racing to complete the sequencing of the human genome have failed, dashing hopes of collaboration and leaving a trail of recriminations.

The two teams — the publicly funded Human Genome Project (HGP) and the private company Celera — had been discussing forming an alliance to pool their resources, but, as in a previous attempt to collaborate (see Nature 397, 93; 1999), they stumbled primarily over the issue of public access to data.

Although the HGP is opposed in principle to any kind of privileged access to data, it had offered a compromise allowing Celera some rights to any data generated as a collaboration for twelve months, in order to speed up completion of the sequence.

Celera's compromise was that the consensus genome could be released publicly at the time of completion, but only on the condition that the data could not be used to compete with Celera's position as a database provider for three to five years. All researchers would have unrestricted use of primary data, says a Celera spokesman, but data merged from the HGP and Celera projects and assembled by Celera “should not be given to those competing with us in the database business for a period of time”.

Both sides are now bristling with anger. In frustration, the UK's Wellcome Trust, a major participant in the HGP, released last weekend a confidential letter sent to Celera by the HGP, numbering the issues it says were hampering talks and asking for a response before 6 March 2000.

The letter — whose signatories include Francis Collins, director of the US National Human Genome Research Institute and Harold Varmus, director of the US National Institutes of Health — outlines the HGP's concerns about Celera's desire to have its commercial rights extended to research applications. These include the construction of genome chips, large primer sets, or applications to proteomics and analysis of regulatory sequences.

“Our position is that we would do nothing in collaboration that would affect our ability to develop the intellectual property on our discoveries,” says Paul Gilman, director of policy planning at Celera.

The letter says that the HGP is troubled by what it perceives as Celera's intention to publish under its own name merged data in a peer-reviewed journal. “We would give appropriate acknowledgement or credit for [HGP scientists'] contributions — if they chose not to collaborate,” says Gilman.

Michael Morgan, chief executive of the Wellcome Trust's genome campus, says he felt forced to release the letter in advance of the response deadline because the Trust had been “frustrated” by Celera's lack of response. “Negotiations seemed to be going nowhere,” he says. At the same time, Celera was making optimistic statements in public, expressing hope that a collaboration would get off the ground, says Morgan.

Celera counters that the early release of the letter was neither ethical nor “in the spirit of co-operation”. This week Venter told The Washington Post that releasing the letter was “a low-life thing to do”. Meanwhile, John Sulston, director of the Sanger sequencing centre, told the BBC Today programme that Celera's taking public data and selling it with their own amounted to something of a “con-job”.

Hopes are now dwindling that the groups could work together. Morgan says this is not the end of the idea but warns that the possibility for collaboration is dwindling. “The train is rapidly leaving the station,” he says.

Gilman declined to say where the collaboration stands “until we have heard directly from the National Institutes of Health or the Wellcome Trust”.