Contents
From February 25 to 28, 1996, the heads of the major sequencing centers meet at the Princess Hotel in Southampton, Bermuda. They are all working toward the same goal: deciphering the human genome. At the same time, they are competitors for money, fame and publications. Over these days they reach an agreement that breaks with a long-standing scientific convention. An example of the Team Principle of External Relationships.
Twenty Labs, One Goal
Situation
The Human Genome Project begins in 1990. It is planned to run for 15 years, with a budget of about $3 billion. The first head of the program at the US National Institutes of Health is James Watson; Francis Collins succeeds him in 1993. At least 20 institutions in six countries take part: the US, the UK, Japan, France, Germany and China.
Connection to the team principle
The Team Principle of External Relationships says: A team that only looks inward loses what it needs from outside. Each of these labs was a team in its own right, with its own people, its own funders and its own ambitions. The task was too big for any one of them. Had they worked in isolation, they would have worked against one another: the same stretches sequenced twice, errors undetected, results left in a drawer for years.
Bermuda, February 1996
Situation
The meeting is convened by the US National Institutes of Health and Britain’s Wellcome Trust, which together pay for about 90% of the sequencing. The driving forces are John Sulston and Robert Waterston, who know the approach from their research on the roundworm, along with Francis Collins and Michael Morgan of the Wellcome Trust. The agreement: contiguous sequences of about 1,000 to 2,000 bases or more are automatically released into public databases, ideally within 24 hours. Finished stretches are published immediately. The entire sequence is to remain freely accessible. In addition, the centers announce which stretches they are working on, so that no one duplicates work.
Connection to the team principle
Until then, researchers released their data when their own paper came out. The Bermuda rule turns this around. Every team shares its data with the outside world every day – and every day gets back what all the others have found. Here, external relationships are not an extra for quiet times. They are built into the workflow. No one has to decide afresh each morning to be open.
Openness Came at a Price
Situation
The agreement did not come about by itself. Two more meetings in Bermuda follow in 1997 and 1998, and in early 1998 the major funders formally adopt the rule. They enforce it through their financial power. Germany, Japan and France had different national rules on data access and had to make exceptions to stay in the consortium. The historians Kathryn Maxson Jones, Rachel Ankeny and Robert Cook-Deegan show that daily release did not serve an ideal alone. It helped in very practical ways with quality control and with coordination between the centers.
Connection to the team principle
The story should not be romanticized. Openness was also achieved through pressure, and in practice it was flexible. That is exactly what makes it instructive. Teams rarely open up out of conviction alone. It takes someone with clout who wants the rule, and it takes a benefit that every team feels in its daily work. In 1999, studying 120 development projects at an electronics company, Morten T. Hansen showed that weak ties to other units help teams find useful knowledge – but that transferring complex knowledge requires strong ties. The Bermuda meetings provided both: a database for everyone, and leaders who sat down together in person year after year.
The Competitor and the Finish Line
Situation
In 1998 Craig Venter founds the company Celera as a private competitor, with a budget of about $300 million. Celera releases data less often and with restrictions. On June 26, 2000, Bill Clinton and Tony Blair jointly announce the rough draft of the genome. In February 2001 the papers appear in “Nature” and “Science.” On April 14, 2003, the project is declared complete – two years ahead of schedule and, at about $2.7 billion, under budget. “Complete” at the time meant about 92%; a gap-free sequence was not available until 2022.
Connection to the team principle
How much the race with Celera contributed to the pace is disputed. What is certain is that a consortium of twenty institutions in six countries held together for years and finished earlier than planned. The Bermuda rule was not the only reason. But it turned twenty teams with interests of their own into a network in which each one thrived on the work of the others.
Key takeaways
Key Lessons
- Build exchange into your workflow. A fixed rule (“Interim results go to the neighboring teams every Friday”) works better than an appeal for openness.
- Share early, not just the finished product. Showing half-finished work takes some nerve, but it saves duplicated effort and brings errors to light.
- Tell others what your team is working on. That one piece of information alone keeps two teams from doing the same thing.
- Make sure there is a tangible benefit. Openness lasts when every team gets something back every day.
- Meet the people behind the data. Complex knowledge travels through relationships, not through repositories.
Summary
In the Human Genome Project, at least twenty institutions from six countries worked on a task too big for any one of them. With the Bermuda Principles of 1996, they committed to releasing their data, ideally within a day. The rule was enforced partly through pressure from the funders, and it served the very practical ends of coordination and quality. It shows the Team Principle of External Relationships as a procedure: a team that only looks inward loses what it needs from outside – those who give every day get something back every day.