I WAS STAYING IN THE PENTHOUSE SUITE IN THE CLAREMONT HOTEL in Berkeley, the white tower that you can see from all over San Francisco Bay. It has a private staircase up onto the roof, where you can stand and see the whole of Berkeley spread out below you. I was in Berkeley as part of a National Institutes of Health team that was reviewing the progress of Gerry Rubin’s lab. Since leaving the LMB Gerry had made his name in Drosophila molecular genetics, and was now beginning to sequence the fly genome. Site visits didn’t normally entail such extravagant accommodation, but on this occasion the Claremont was short of rooms and so I got this splendid suite as part of the NIH contract. It provided an appropriately sumptuous backdrop to what followed, as if the scene had been set up by a film producer.

That evening in January 1992 Frederick Bourke, a wealthy investor who had made a fortune in leather goods, came up to see me. It was our third meeting, and things were not going the way he expected. At one point he remarked that ‘dealing with scientists is like being a sheepdog rounding up sheep.’ He wanted Bob Waterston and me to head the commercial sequencing organization

he was hoping to establish in Seattle. As on previous occasions I was non-committal, but this time he probably saw in my eyes that I wasn’t going to accept. He said ‘John, I do hope that this isn’t going to do you any damage.’ I said it would do me no damage at all. There was no harm in being known as somebody who was wanted in America.

After Bourke had gone, my daughter Ingrid and her boyfriend Paul Pavlidis came over to the hotel with friends. Ingrid had come to Berkeley in 1990 to do a Ph.D. in developmental biology. She had been so eager to go to California, I used to tease her about being imprinted on the land of her birth. We went up on the roof and smoked a little, as one does in Berkeley, and looked out over the bay. I remember thinking that it was another significant moment. But I didn’t know that within months, and as a direct consequence of this slightly surreal episode, Jim Watson would have lost his job heading the NIH genome program and I would be committed—that prison door again—to heading the laboratory that would ultimately complete the sequence of one-third of the human genome.

Given our commitment to the public domain, the fact that Bob and I were even considering heading up a commercial organization needs a bit of explanation. It was all because of the success of the pilot worm sequencing project, launched in Jim’s office two and a half years previously at Cold Spring Harbor. By the end of the second year it was clear that we were going to meet the targets set for the three-year pilot project. I had never doubted that we should then scale up from doing 3 percent of the worm genome to doing the other 97 percent, and Bob was just as confident. But we were very unclear in our own minds whether either of the funding agencies on which we depended was going to come up with the money we would need. Our doubts arose partly because of the flak we were getting at this stage from those who doubted the utility of the worm genome, but mainly because the sums involved were going to be so

large: we had slashed the cost per base, but we would still need in total ten times as much as we had had before. We weren’t very confident that our funders would think we were worth that kind of expense, so we were very open to exploring other possibilities. We had never before had this sort of responsibility, both to our project and to the teams working on it. Bob and I were both worried about it. We certainly planned to put in further grant proposals; we were just concerned that we might not get the money. For the next stage there would be no more funds coming to the U.K. from the United States government. Jim made this absolutely clear: for the pilot phase he could do a deal, for the production phase, no. If NIH was going to put in half, then the MRC had to put in half. And now the MRC was really challenged: they would have to find about £10 million.

Then, in 1991, with a year to go before our pilot funding ran out (Bob’s had come later, so he had a little longer), Lee Hood phoned up out of the blue and said, ‘I have a proposition for you. I want to start a sequencing organization, and I want you and Bob to come and lead it.’ Lee is a very remarkable guy, and a great networker. The fact that we were sequencing so efficiently owed a great deal to the ingenuity of his CalTech lab in starting the development of the automated sequencers, followed by the founding of ABI under Mike Hunkapiller to turn them into a commercial product. Lee had just agreed to move to the University of Washington, Seattle, to start an academic biotechnology department with money from Bill Gates. Now, he told us, he was proposing to start a commercial sequencing operation alongside the department, and he had persuaded Rick Bourke to put up the capital. He wasn’t offering us academic positions, but we would be somehow affiliated to the university, and he implied that we would be able to complete the worm genome sequence.

If it had been anyone but Lee, we might not even have considered it. But he was someone we respected, who was running a good

academic lab. We were intrigued enough—and doubtful enough about our chances of getting funded by our government funding agencies, or indeed anyone else—to fly over there and have a couple of meetings with them all.

Initially, Bob and I were much taken with the idea. We would have been in the same city, working side by side. ‘That was a very attractive part of it,’ says Bob. ‘Lee Hood was going to be there, with the technology push that he had. He was also trying to attract Maynard Olson to his group [Maynard moved there soon afterwards], and you could see the attraction of setting up a real local powerhouse.’ We were seriously trying to think, ‘What does Lee want? Do we want to do this?’ Bourke was offering us salaries that were enormous by my standards, and of course stock options; and, more to the point, if we wanted to get into large-scale sequencing this might be one way to do it. And then, little by little, we began to realize that sequencing the worm wasn’t on Bourke’s agenda at all. He began to complain about how ‘all scientists came with baggage.’ And when we began actually to write down what we were going to get out of the deal, he started saying things like, ‘How many machines can you bring with you?’ More seriously, we differed on the question of releasing the sequence data into the public domain. Bourke’s initial idea was to patent the sequence. We said that wasn’t necessary, we could just introduce a delay of a few months before releasing the sequence that would give a research team time to find genuinely useful regions. He didn’t buy it.

It gradually became apparent that we were simply going to be scientific directors of a commercial sequencing organization, and there was less and less possibility that we would get the worm done at all. It was quite clear that all we would get would be premises—which we already had; if we wanted to do any of our own research, we would have to go out and get grants in the usual way. And with both of us in the United States, we would have to go after NIH money jointly; we would lose the international collaboration, with

the MRC as equal partner.

By the time of my meeting with Bourke in the Claremont Hotel, we were close to deciding against the scheme. Soon afterwards we set up a conference call and Bob and I told him it wasn’t going to happen. I was never in any difficulty about the decision. One has one life and one life only, and one does what one really wants to do. I don’t see the point of making money for its own sake.

Meanwhile Jim Watson got wind of the negotiations with Lee Hood and Rick Bourke, and as far as he was concerned, Bourke was trying to pinch his project.

It was true that, for a variety of reasons, the other pilot projects that had been funded in the first round of genome project grants, on E. coli, yeast, Drosophila and parts of some human chromosomes, were not producing anything like as much sequence: the worm was, in effect, the only game in town. Jim arranged to meet Bourke, and was incensed to find him bent, as Jim saw it, on destroying the NIH genome program for a private company. As well as wanting to retain the most successful sequencing teams, Jim was very anxious not to lose the international aspect of the sequencing project, which at the time was basically the worm genomics group and the MRC.

Jim immediately began to contact everyone he could think of to prevent us from joining Bourke. But first he rang Bob Waterston. At

this point, in mid-January 1992, we had still not finally made up our minds. After talking to Jim, Bob realized that we could turn the situation to our advantage—‘use it to open the spigot a little bit,’ as he puts it. He wrote to Jim to let him know in uncompromising language that our basic problem was lack of funds. While the NIH was apparently sitting on its hands, new genome efforts were beginning in France and the United States with private or charitable support. ‘We did not start this project to be outdone by second-rate players with more resources,’ wrote Bob. But he emphasized our commitment to keeping genome work in the public domain. ‘We, of course, have some remaining misgivings about a venture so foreign to what either of us has done before…the genome project cannot prosper if it is viewed simply as a way of making its creators rich.’

The same day that Jim called Bob, he also spoke to Aaron Klug at the LMB, poured out his ‘annoyance and rage’ at Bourke’s venture, and told him that he would be in England at the end of January on his way to the World Economic Forum in Davos, Switzerland. On the way he called in to see Dai Rees, secretary of the MRC, to tell him how concerned he was about the possibility of my being lured to the United States. Then he came over to Cambridge to talk to me. I confirmed what he had already heard from Bob—that we had considered Bourke’s offer only because we couldn’t be sure we would get the money to complete our project. But by this time we were close to saying no, and he left reassured that we weren’t on the point of deserting him.

Jim probably didn’t realize that, by rushing around making a stink on our behalf, he was locking the stable door when the horse was still firmly inside and had bolted itself in. I don’t know why we hadn’t taken him into our confidence much earlier. As it was, his intervention added fuel to a fire that had already been lit underneath him at the NIH.

James Wyngaarden had been succeeded as head of the agency by Bernardine Healy, a career scientific administrator. She and Jim did

not see eye to eye. The main bone of contention between them at this time was the patenting of gene sequences. The previous summer, the NIH had filed patents on several hundred fragments of genes, known as expressed sequence tags or ESTs (see p. 105). The ESTs had been generated in the lab of Craig Venter, in the course of research that was independent of Watson’s genome center. ESTs are convenient labels that help to identify genes, but do not in themselves tell you anything about the function of a gene (unless you can match them to genes in the same or other species that already have known functions).

Patents (or so I had always believed) are designed to protect inventions. There are three essential criteria for an invention: it has to be novel (no-one has published the idea before), useful (in that it could be developed for commercial or other uses) and non-obvious. The ESTs met none of these criteria. There was no ‘invention’ involved in finding them, so how could they be patentable? Yet the 1991 patent application claimed exclusive rights not only to the ESTs, but to the whole genes they represented and even the proteins encoded by these genes. It was crazy. But Healy had arrived determined to push the commercial development of scientific discoveries. There had been no proper debate, nationally or internationally, about where the line should be drawn on patents for genetic discoveries, but the NIH lawyers decided to play safe by putting in patent applications on Craig’s ESTs. That way, if scientists in other countries started patenting genes, they would have a prior claim.

The decision to apply for the patents involved a number of different branches of the NIH bureaucracy, but the genome research center and its head, Jim Watson, were told it was happening rather than asked for their opinion. When Craig announced at a public briefing on genome research for the United States Senator Pete Domenici in July 1991 that the patents had been filed, Watson burst out that the move was ‘sheer lunacy’, and said he would be ‘horrified’ if it were true that random bits of sequence could be patented.

He argued that there was no invention involved, asserting that the automation of sequencing meant that the work could be done by ‘virtually any monkey’—a typically unguarded remark that was probably more insulting to Craig and his colleagues than was intended. Jim was extremely concerned that premature patenting of sequences whose function was unknown could undermine the delicate structure of national and international collaboration that he saw as essential if the benefits of the genome project were to be fully realized. Healy, on the other hand, believed that without patents on the sequences potential licensees from United States industry would have no interest in pursuing commercial development of the discoveries. It was not a matter on which the two of them could agree, and their dispute continued through the medium of public pronouncements on either side rather than face-to-face discussion.

Watson eventually agreed not to criticize NIH policy publicly, but others in the academic community could speak for themselves. The scientific advisory committee on the Human Genome Project, chaired by Paul Berg from Stanford University, said it was ‘unanimous in deploring the decision to seek such patents…we believe such claims are inappropriate and deleterious to science.’ Like Jim, the committee feared that the claims would set off an international ‘patent race’ that would destroy the collaborative framework of the genome project. HUGO had already made a similar statement and Walter Bodmer, now HUGO’s president, confirmed that if the United States agency persisted with its patents then the U.K. would have no alternative but to follow suit. Despite these protests, in February 1992 the NIH added a further 2,375 ESTs to the patent application, although this time they dropped their claim to the proteins encoded by the genes they represented. Their action set off a fresh round of criticism. Berg spoke for many when he said, ‘It makes a mockery of what most people feel is the right way to do the Genome Project.’

Into this volatile atmosphere came a letter from Bourke to Healy,

backed up by a phone call from Lee Hood, accusing Jim of acting against the interests of American industry by interfering in Bourke’s attempt to recruit Bob and me. ‘I was not anti-American,’ says Jim now, ‘but I was certainly anti-Bourke because I didn’t want the one success that we had broken up and in the hands of private industry.’ Healy used Bourke’s concerns as an excuse to set up an investigation into Jim’s financial holdings on the grounds of suspected conflict of interest. Nothing was found that he had not previously declared, but Healy never publicly exonerated him. Jim felt he was left with no option. He resigned in April 1992, but today he is in no doubt that he was effectively fired.

Jim returned to Cold Spring Harbor declaring his continued support for the project under whoever turned out to be his successor. But the HGP had lost a hugely valuable asset—a bold and single-minded leader who would battle publicly for what he believed in, and one who had made the project truly international. At least Jim had established the principle that the project should be led by a scientist and not a career administrator. A year after his resignation Healy appointed Francis Collins, the University of Michigan geneticist who had been involved in finding several disease genes including that for cystic fibrosis, to succeed Jim. In the midst of a productive phase of his own research career, Francis was a reluctant recruit, but agreed to move to the NIH campus at Bethesda if the job could include facilities for his team and other new genetics investigators to pursue in-house research. The National Center for Human Genome Research later became the National Human Genome

Research Institute (NHGRI), which both conducted its own research and dispensed funds to other genome centers. Francis, a tall, spare figure and a devout Christian, was to remain in the hot seat—the NHGRI became the de facto coordinating center for the international genome project—through to the present day, deploying his gentle Southern charm in an effort to build consensus among the often hostile factions that had an interest in the human genome.

The controversy about the EST patents rumbled on for another two years. The United States Patent Office roundly rejected the first application in August 1992, but the NIH appealed against the rejection and applied for patents on another 4,448 ESTs. After the election of Bill Clinton as United States President in November 1992 Healy resigned (the NIH head is a political appointment and she was a Republican appointee). Her replacement, the highly respected cancer researcher and Nobel prizewinner Harold Varmus, decided not to pursue the patent issue any further and withdrew all outstanding applications in early 1994. The question of what constitutes an acceptable patent on genetic material is still not fully resolved, and arguments sway to and fro. The United States Patent Office in particular has granted thousands of patents on gene sequences, but anyone with the funds to challenge these patents in the courts can do so. Not until 2000 did the Patent Office produce a set of guidelines that tightened up the definition of ‘utility’ to prevent people giving uses as vague as ‘a gene probe’ in their applications. It is still permissible to patent a gene sequence as long as you can show how it might be used to diagnose diseases, for example. In the end the issues are being decided not on principled grounds, but according to which side has the most money to spend on lawyers. One of the aims of the Human Genome Project has been to ‘raise the bar’ by making as much genome information as possible universally available in the public domain and therefore unpatentable.

How much the Bourke episode contributed to Jim’s departure from

his key position in the genome project is hard to tell—the patent issue was probably the more serious source of conflict between him and Healy. But although from his own point of view Jim risked more than he needed to in order to keep Bob and me where we were, his intervention made a material difference to the fortunes of both the Cambridge and the St. Louis worm sequencing groups. Jim made the MRC realize how much the U.K. stood to lose if it failed to fund genome sequencing at a substantial level. After his visit, Aaron Klug immediately got Dai Rees’s permission to contact Bridget Ogilvie, the recently appointed director of the Wellcome Trust.

The Trust is a charity that was set up by Sir Henry Wellcome on his death in 1936. He placed the entire share capital of his successful pharmaceutical company, the Wellcome Foundation (later Wellcome plc, then Glaxo Wellcome, now absorbed into Glaxo SmithKline) in the hands of trustees, and directed them to use the income from the shares to fund scientific research for the benefit of human and animal health. In the mid-1980s, following concern that the company’s performance had reached a plateau, the trustees began to sell off their shares in order to widen the asset base and safeguard the substantial contributions they were making to research. At about the same time, there was a dramatic upturn in the company’s fortunes following the introduction of the anti-HIV drug AZT. After the Trust sold a second tranche of shares in 1992 it became the wealthiest medical research charity in the world. Bridget Ogilvie, an Australian scientist who had worked for the Trust since 1979 after a productive research career investigating immune responses to parasitic nematodes, became its director in October 1991. Within a year her budget had more than doubled, from £91 million to £200 million (today it is more than twice this figure), all of which she had to spend. ‘To spend money on that scale sensibly is not easy!’ she says.

The MRC’s approach to the Wellcome Trust could not have been

more timely. Although relations between the Trust and the MRC had been strained in the past, Bridget (whose own research career had been conducted mainly in an MRC institute) was enthusiastic about entering into some kind of partnership.

Very quickly it became apparent that the Trust was not interested in subsidizing our work at the LMB, but was prepared to think about spending very considerable sums on a much larger genome sequencing effort that would not only provide space for the worm sequencing group, but would also embrace human DNA. The MRC was put on its mettle—it wasn’t going to allow the Wellcome Trust to steal away its project. In June the following year it came through with a grant of £10 million over five years to complete the worm sequence—a wonderful vote of confidence in the project that meant so much to me.

The year of 1992 was completely insane. It was one of those moments in one’s life when one feels swept along like a leaf on the stream. First there was Bourke, and a burst of comment about the episode in the British press, which perhaps helped to propel succeeding events. Then I got sucked into negotiations with the Wellcome Trust for building a new sequencing center, of which I would be director, to make an attack on the human genome. In the spring Aaron Klug dragged me down to London to a meeting with the Wellcome Trust’s Genetics Advisory Group. The group had been set up only the year before to develop a strategy for the Trust on genome research. At Aaron’s instigation, I had hastily written a four-page briefing document, plus a few tables on the running costs, which I put at just over £10 million for the first two years. In it I described our progress on the worm genome and set out how this

experience could be transferred to the human. I concluded, ‘Surely now is the time to begin sequencing on a large scale, rather than, as has been argued, waiting upon the emergence of still better instrumentation and ingenious genetic tricks…To begin now gives Britain the opportunity to show the way.’

In the meeting Aaron adroitly made the case that large-scale sequencing, or megabase sequencing as we called it, was going to change things totally in the search for disease genes. He took the example of the gene for Huntington’s disease, which had been linked to a chromosome almost a decade before but was proving extremely difficult to locate definitively. Trying to get ever closer to the gene through traditional genetics, looking for linkage with other markers, took a lot of time and a certain amount of luck. On the basis of what we were doing in the worm, Aaron argued that once you had got within a few million bases you could stop doing linkage studies and sequence selected mapped clones instead. ‘That was the argument,’ he says; ‘that you would be able to get results on the way—even before you had the whole sequence you would be able to home in on the gene of interest.’

Everything began to fall into place with astonishing speed. In March 1992 Bridget Ogilvie appointed one of the senior Wellcome Trust administrators, Michael Morgan, to look into the viability of a Trust sequencing initiative. Meanwhile we had to develop my briefing note into a formal proposal to cover the first five years. We were talking about a highly organized operation several times the size of our existing worm sequencing group—bigger, indeed, than any other sequencing center in the world at the time. The proposal was for a grant of between £40 million and £50 million. (And to think that less than three years before I had gone weak at the knees at the sight of a £1 million grant!) Considering the sum of money at stake, our proposal was really a very tiny document. It included a table that I made up almost as a joke one day, that showed how in five years we were going to finish the worm, and complete

the sequence of baker’s yeast (Saccharomyces cerevisiae) alongside the international consortium that had recently started, and sequence the first 40 million bases of the human. Actually, we very nearly did it: we were just about six months late all through, which was the difference between the optimistic nominal start date and our actual move to the new site. Anyway, it worked. The governors considered the proposal, and agreed to fund it, just like that. ‘One of the pressures was to do it quickly,’ says Bridget Ogilvie.

In the summer of 1992, anticipating (correctly) that there wouldn’t be another proper holiday for a long time, Daphne and I took off for Glacier Park in Montana. Instead of going direct, we flew to Seattle and hired a car there. There was a little nostalgia in looking around Seattle on the way, this beautiful place where we had almost gone to live. Glacier was perfect, with its lakes and mountains and long footpaths and bears. Ingrid came to join us halfway through, riding the long train that snakes slowly up from the coast, and together we walked the two-day trail over the continental divide. It was a welcome interlude in a chain of events that was sweeping me along, to what end I could not begin to predict.

During the holiday I put in one or two calls to Michael Morgan, to see how things were going. One of our first tasks was to find a building to put the new center in. Over the phone he told me about various possibilities that had come up, including a disused chicken research institute—he joked that the big shed looked like a possibility for the sequencing machines, even though it was too low for a person to stand upright. When I got back from holiday Michael and I spent a few weeks going round looking at office blocks in Cambridge, without much success. We also briefly considered Edinburgh; it was a growth area for hi-tech industry, and would have been a wonderful place to go—like Seattle, with hills and water. But regretfully I turned my back on such frivolous considerations: it would have delayed us to move so far because not everyone

would be able to go, and it seemed wise for reasons of continuity to stay in touch with our base in Cambridge.

Then we heard from Michael Fuller, the lab superintendent at the LMB, that a property developer was letting an estate in the village of Hinxton, about nine miles south of Cambridge, which had previously belonged to the metallurgy and engineering firm Tube Investments. The estate included the rather run-down Hinxton Hall, a small stately home dating from the eighteenth century, plus 55 acres of parkland and a range of brick laboratory buildings that had been built as a research center for TI. The developer had intended to knock down the labs and build a business park, and had planning consent for 125,000 square feet of new buildings. But finding no takers for his plan in the early 1990s property slump, he put the whole estate back on the market. We went to have a look, and were charmed right away. The estate is set among open fields, and the river Cam runs along the boundary on its way to Cambridge. Michael and I saw that the labs could be converted into a genome research facility in a matter of months. There was ample space to get going, though the situation out of town and 8 miles from the LMB caused some heart-searching. The Trust agreed to let us take a one-year lease on the estate and carry out the conversion. But the refurbishment of the Tube Investments buildings had barely started when another major development took place that materially affected the future of the project.

During the winter of 1992–3 I was on a committee talking about the future of the DNA Data Library of the European Molecular Biology Laboratory, an international research organization funded by a consortium of countries. The Data Library, one of three international repositories for sequence data, along with GenBank and the DNA Data Bank of Japan, was still housed in the European Molecular Biology Laboratory’s headquarters in Heidelberg. The committee was discussing the idea of developing not just a growing sequence repository but a European Bioinformatics Institute. The

institute would store data but also develop the tools to do research on it in a computational way. The Germans assumed the new institute would also be in Heidelberg; they had planning permission, and were all set to build it there.

Michael Ashburner, the fly geneticist, was also on this committee. He took it into his head to persuade the Data Library’s director Graham Cameron that we should get the whole thing to move to the U.K., and I backed him. The Wellcome Trust and MRC were enthusiastic about the idea, and they moved quickly to establish the principle that there should be an open competition for the institute, rather than letting it go to Germany by default. We then began to put a bid together. Immediately we hit a snag: neither Wellcome nor the MRC owned a suitable site. Michael Morgan proposed Hinxton, although the Trust did not actually own it at the time. The Wellcome Trust governors solved that problem by agreeing, at very short notice, to buy the whole estate. So out of the blue these cheeky Brits put in a proposal to host the European Bioinformatics Institute and won, on the basis that the institute should move in next door to what was going to be Europe’s largest sequencing center.

That coup determined us to stay at Hinxton. Although the idea of the lab being there on a temporary basis was established, we had not previously decided that it was going to be permanent. The estate cost ‘two or three million’ pounds to buy; it was a perfectly good investment, with the property market in such a state of decline. But winning the bid for the bioinformatics institute prompted Wellcome to make a much bigger investment in the site. By the time we moved to the refurbished labs in April 1993, there were already plans to replace them with a sleek, modern laboratory surrounded with pools and lawns, and to restore Hinxton Hall as a conference center. The institute would go up next door, and in 1994 the MRC also moved its Human Genome Mapping Project Resource Centre to the site, now known as the Wellcome Trust Genome Campus.

We decided to name our new lab the Sanger Centre, in honor of

Fred Sanger, who had first made DNA sequencing a practical possibility. It fell to me to phone Fred up and ask his permission to name the place after him. I dialled his number in some trepidation, not knowing what his reaction would be, for Fred famously shuns publicity. He accepted immediately, but on one condition. ‘It had better be good,’ he said, and once again I heard the door closing, as it had at Syosset. In October 1993, a few months after we had moved in, we held an enjoyable opening ceremony at which Fred himself unveiled a commemorative plaque and we presented him with a security pass to the building so that he could visit whenever he liked.

Money and buildings were all very well, but the success of the Sanger Centre was going to depend on the people we found to staff it. From the moment it was first mooted, I realized that I would need an excellent administrator, never having run anything like this before—though it’s not quite true to say I had never held any position of responsibility other than running the worm sequencing team. Not long before, Aaron had made me head of a new section at the LMB called Genome Studies. When Sydney left in 1986, he left behind a loose grouping called the Director’s Division, and at least as a stopgap Aaron felt the best thing was to keep it together as an entity with me as head. ‘It seemed very clear to me that although [John] eschewed responsibility he was actually taking it on willy nilly,’ says Aaron. But even so, heading Genome Studies, which consisted for the most part of independent scientists, was not going to be the same as running a genome sequencing operation employing dozens of people. And, although I probably didn’t quite realize it at the time, I was going to be directly responsible for the day-to-day budget in a way that I’d never previously experienced. I was going to need someone with real administrative skills.

Jane Rogers had moved from scientific research in Cambridge to scientific administration at the MRC head office in London. But she still lived in Cambridge and, with a young son to look after, found

the commuting difficult. So she had asked if there was any possibility of being transferred back to Cambridge, and they sent her to talk to me. Alan Coulson and I interviewed her in my sitting room at home, an experience she found ‘bizarre…They plied me with sherry and asked if I could convert an office block into a lab,’ she recalls. It seemed right straight away, and was to be the first of many interviews I would share with Jane, though thereafter we would be on the same side. (I find that usually after five minutes you know whether you can work with someone or not; from then on it’s either real fun as you explore one another’s thoughts or else you’re trying to get out of the room as fast as possible.) I told the MRC that we were keen to employ Jane, and around the middle of 1992 she was seconded to us for two or three days a week, initially still in the lab at the LMB. The first thing we had to do was to write the formal proposal to the Trust that would lay the foundations for our as yet unnamed genome center. I tried to write something, dropping drafts through Jane’s door as I cycled home in the evening for her to work on by the morning. The only grant proposal I had ever written before was the worm sequencing proposal, and Bob had had a lot to do with that, but through her experience at MRC head office Jane knew exactly what funding bodies were looking for. ‘I knew how to write down what was needed and make it look plausible,’ she says, and she played a big part in sorting out the budget pages so that they added up to the right amount.

The next question—of course, all these discussions were going on in parallel—was which scientists would move to the new center. It was to be a joint enterprise between the Wellcome Trust and the MRC, but it was clear from the start that in financial terms the partnership was unequal. What the MRC could bring to the table, in addition to their funding of the worm sequence, was the expertise they had nurtured at the LMB. Alan and I and our worm sequencing team would go, obviously. And Bart Barrell, who had sequenced more genomes than anyone and was now working on

yeast, jumped at the chance to come.

Then we needed people who could manage the bioinformatics side, writing the software to manage the flow of information, analyze it and display the results. Initially we expected that Rodger Staden would join us. For years he had written the sequencing software used at LMB, at first on his own and then leading a growing group, and the Staden package is widely used. He eventually decided against leaving the LMB, but we have continued to collaborate closely. Richard Durbin, who had written ACeDB with Jean Thierry-Mieg to handle the worm sequence, did move across to head the new center’s bioinformatics group, and later he became deputy director.

If we were going to work on human DNA, we needed someone who knew something about it. All DNA is alike in that it all consists of the same As, Ts, Cs and Gs. But compared with the worm, for example, human DNA presents particular difficulties for the sequencer. One problem is the high proportion—more than 98 percent—of non-coding regions; the other is the fact that as much as half of the non-coding DNA is in the form of repeats. These repeats make assembling the DNA fragments read by the machines a bit like doing a jigsaw puzzle in which half the pieces depict either grass or sky. And, apart from the technical difficulties of reading human DNA, there was an establishment of human geneticists all with interests in particular human disease genes, a world about which I

knew little. We needed someone with a foot in the human genetics camp, but who was also interested in human DNA for its own sake.

Right from the start I had in mind David Bentley, who at the time was working in the genetics department at Guy’s Hospital in London. David had worked on a detailed analysis of the mutations in the Factor IX gene, a gene on the X chromosome that causes a form of hemophilia. But he was also interested in mapping larger genomic regions, and clearly understood that in the long run genomic mapping and sequencing was going to make the search for disease genes a whole lot easier.

Earlier, David had formed his own collaboration with Ian Dunham and Eric Green in Maynard Olson’s group at St. Louis, putting together a YAC map of the muscular dystrophy region of the X chromosome. After this Ian, a British scientist who had been doing a post-doc with Maynard, came back and joined David at Guy’s. They embarked on a map of the whole of chromosome 22, one of the smallest human chromosomes.

Throughout this period David used to come and visit Alan and me at the LMB to talk about fingerprinting, and Richard to talk about ACeDB, which Richard had adapted to handle human data. ‘I always enjoyed going up there,’ he says. ‘It was a breath of fresh air to talk technology and not be worried about genetic targets.’ So when I began to think about who we needed to look after the human side of things, David seemed the obvious candidate. Inadvertently, I picked a rather awkward moment to approach him. He was in a meeting with his head of department, Martin Bobrow, when my call

came through. Oblivious to the delicacy of his situation, I said, ‘I’ll come straight to the point. How would you like to come and join me in setting up an institute?’ David nearly fell off his chair with surprise, and was left struggling to find a suitably diplomatic reply. But very soon he accepted.

David brought with him his own project on the X chromosome, as well as Ian Dunham and the chromosome 22 mapping effort, so that from the start we had flagship projects on the human genome.

I thought the team of myself, Alan, Jane, Richard, Bart and David would be all we needed to manage the various aspects of the new center’s operations. But the Wellcome Trust required us to have a chartered accountant to look after the financial and legal side. That was a foreign move for all of us—the MRC would never have made such an appointment, because these matters are handled by head office. But, because we were funded jointly by the Wellcome Trust and the MRC, a management company called Genome Research Limited was to be set up and would require a qualified company secretary. So a smart-looking advertisement went out in The Times, and the Trust recruited Murray Cairns, a genial Scot who had been made redundant after many years in management with Bass, as a result of the restructuring in the brewing industry. In order to make clear the respective responsibilities of Jane and Murray, she was called the Scientific Administrator and he the Head of Corporate Services. (Jane later handed over the scientific administration to Christine Rees, and became Head of Sequencing.) We worried at first that Murray might cramp our style by being too bureaucratic,

and indeed I heard much later that the Trust hoped he would keep an eye on us as their watchdog. But despite his initial indoctrination he quickly ‘went native’ and became a doughty champion on our behalf against the bean-counting element at the Trust’s headquarters in Euston Road. He also made us aware of our shortcomings as managers, and gradually helped us to see that, in an operation that was both a factory and a laboratory, it was essential to acquire management skills. Over the years Murray and I have spent more and more time together (particularly in the Red Lion at Hinxton) and learned a lot from each other.

This, then, was the team—Alan, Richard, Jane, David, Bart, Murray and me—that became the board of management (known as the BoM) of the new center. As I write, Murray and I are the only ones of the original team who have retired—all the others are still in place, a testament to their commitment and the excellent team spirit they developed.

In the first year after our move to Hinxton the staff grew from fifteen to eighty, and it continued to grow as the years went by. Jane and I undertook the initial recruiting of the staff we needed to run the sequencing. We ran our first advertisements in the Cambridge Evening News, and got people ranging from school leavers to a graduate in philosophy. According to Jane I had shocked the LMB by recruiting people who had no academic background. She accuses me of having favored women who’d worked as barmaids—I was certainly more interested in evidence of practical competence than paper qualifications.

I also had to get used to people leaving. At first I found it hard not to take this personally—how could this colleague not want to see the thing through as I wanted to? Then I realized two things. One was that the organization, as well as the people, was changing and developing very rapidly. A person who made a good fit at one stage was not necessarily a good fit later. The turnover was healthy and allowed evolution. The second thing was that we were

actually training and educating people, just like a university, and so making a practical contribution to the needs of industry—U.K. industry in particular—even though we could not sell them anything.

I learned to share everything with the BoM and to pass on as much as possible; there was so much to be done that you had to delegate part of the planning, let alone the implementation of the plan by other people. I saw my role as that of a chairman. I vaguely assumed that someone would let me know if there was a problem, and didn’t at first see the need for anything more formal in the way of management. But after a bit Murray persuaded us to go for some management training, and that was a real eye-opener. All the members of the BoM went away for the weekend to a hotel in Rutland for an intensive course with someone from a professional management training company. The first thing he did was present us with the results of a survey he had carried out among the hundred or so staff, which revealed that they had a pretty low opinion of our management skills. Some of them were quite unhappy—remarks about piss-ups and breweries were the milder ones. I was fairly abashed; I had no idea it was so bad. Although there was a certain amount of hilarity over the practical team exercises we had to do on the rest of the course, we did realize that we had something to learn. After that weekend we developed a more formal management structure throughout the whole organization, and gradually increased the number of levels in the hierarchy so that there was a career progression for people who did well. The other important outcome was that we realized how valuable it was to get together for a weekend like that, and thereafter we had a retreat every year—not for more management training, but just to discuss the scientific policy of the center and reflect on what we were doing.

The Sanger Centre was a totally different environment from the one in which I had worked until 1989, quietly putting together the worm

cell lineage and then the worm genome map. Indeed, some of my oldest colleagues and friends could not hide their astonishment that I should have ended up in such a position. I remember standing with Bob Horvitz at the window of the old lecture room a year or so after we moved to Hinxton, looking at the enormous hole in the ground that was going to be our new building. ‘John’, he asked me, ‘do you really know what you’re doing?’ I replied that I realized that it was a big change, but that I thought it was important; so I was prepared to do whatever was necessary to make it happen. For myself, I didn’t see it as a permanent change of direction, just as a means to an end. I always believed that once it was up and running I would be able to go back to doing something more sensible. I had forgotten what Alan Coulson had said during that walk in the woods near Heidelberg—you can never go back.

In many respects, things went on as before. Most importantly, the collaboration with St. Louis not only survived the change, but if anything was strengthened. Six months after we got our £10 million from the MRC for the worm sequence, the NIH also came through with funding for a scale-up in Bob Waterston’s lab. Bob is in no doubt that the two events were connected: ‘When it was clear that the Wellcome Trust and the MRC were going to create this larger venue for sequencing over there,’ he says, ‘it made the NIH funders much more receptive over here.’ At the same time he also gained more space, establishing the Washington University Genome Sequencing Center on the fourth floor of an office building that the university had recently acquired.

By the end of 1993 our collaboration on the worm was not only the most productive genome sequencing operation in the world, but the biggest. We had easily exceeded our target of generating 3 megabases of finished sequence in our first three years. On top of that, Bart’s group was churning out yeast sequence, and David was lining up some of our first human cosmids. But as we began to establish our reputation, the first signs had appeared of a belief that

private enterprise could do as well or better than the public project, and could do it quicker and more cheaply. The first in the field was Craig Venter. Frustrated at the lack of support within the NIH for his venture into genome research, he left in July 1992 to set up his own privately funded, non-profit-making genome center, The Institute for Genomic Research (TIGR), in Rockville, Maryland.

Craig had joined one of the in-house NIH labs in 1984, and spent years looking for the gene for the receptor molecule on cell surfaces that recognizes the neurotransmitter adrenaline. But after he acquired one of the first ABI sequencing machines, he focused on a faster way of finding genes by generating expressed sequence tags or ESTs. This technique had been developed by Paul Schimmel and his colleagues at MIT to look for muscle genes as early as 1983. It involved extracting RNA from tissue and using it as a template to make complementary DNA, or cDNA, thereby isolating the protein-coding sequences of genes. Sequencing 150–400 bases at the ends of the cDNAs gives a pair of unique tags for each. Using these ESTs to probe the sequences already deposited in the public databases, you can find out if they represent known genes, if they have similarities with genes from other species, or if they come from previously unknown genes. By the spring of 1991 Craig had completed his first batch of ESTs derived from brain tissue, and he wrote to Jim Watson to propose that, instead of sequencing human chromosomal DNA, with all the difficulties of assembly and interpretation, the genome project might fund him to attack the protein-coding regions first by extending the EST approach. But the genome research center stuck to its policy of beginning by mapping out the whole genome. Although ESTs would later prove to be extremely useful in assisting that mapping effort, the study section that reviewed grant applications declined to support Craig’s proposal.

Craig cannot have improved his chances by publicly positioning the EST strategy as a cheaper and more efficient alternative to genomic sequencing. He published his first paper on his EST work

in Science in June 1991, while he was still at the NIH lab. In an accompanying news article, he described his approach as ‘a bargain by comparison to the human genome project…We can do it for a few million dollars a year, instead of hundreds of millions.’ Although he added that ‘The cDNA approach does not eliminate the need for the Human Genome Project,’ he was clearly trying to bring about a shift in policy through his public pronouncements—a tactic he has deployed frequently since.

We had already been personally affected by Craig’s competitive streak. Some time that same year we heard that his lab was sequencing worm ESTs. I don’t think I’m being paranoid in suggesting that they were trying to compete with us; there had been quite a lot of publicity about our getting the sequencing grant. At some point we held a conference call in which Craig argued forcefully that ESTs were a better way of finding genes, that he’d found a gene that we’d missed and so on. It’s true that the program we used to hunt for genes automatically in the sequence was very elementary in those days, but I was in no doubt that in the long run sequencing the complete genome would be the only way to find all the genes. I saw Craig’s challenge as a threat to what we were doing. If his lab was able to identify a substantial number of worm genes at a time when we had only a few dozen, because that’s all there were in the clones we had sequenced, it would probably not help our case for future funding.

We had started thinking about sequencing our own ESTs in late 1990. Bob thought it would be a good idea to do a batch in parallel with the genomic sequencing; although it wasn’t originally part of our program, we figured we could afford to do it. Chris Martin had just moved from the lab of Marty Chalfie (who was working with the mechanosensory mutants) at Columbia University to Bob’s department at Washington University, and he had made a library of worm cDNAs. Bob persuaded him to let us sequence ESTs from this library, which he copied and sent to us so that we could do half

each, beginning in April 1991. We called our project a ‘survey’ of C. elegans genes—it was not meant to be exhaustive, but to show the range of genes. Our paper came out in early 1992, back-to-back with a paper from Craig’s lab on worm ESTs, in the first volume of Nature Genetics. It was quite useful to have done it, but I also wanted to put down a marker saying that we could do this just as well as anybody else—and now we were going to get on with sequencing the genome, thank you very much.

I saw cDNAs as a means to an end, not the end itself. And they would never give you all the genes. In the news article accompanying his Science paper on human cDNAs, Craig estimated that he could get 80 or 90 percent of the total number: in the same article, I was quoted as saying ‘Eight or nine percent is more like it.’ (This was a joke on my part—I just meant to emphasize that 80–90 percent was optimistic. I learned from that never to joke with reporters; having a sense of humour is not part of their day job.) Equally seriously, cDNAs alone give you no information about the regulatory regions of the genome that turn genes on and off. For complete understanding of the genome, there was only one option, and that was to sequence the whole thing.

Of course, the National Center for Human Genome Research could have regarded the EST work as complementary and funded Craig as well as the other genome centers. But, perhaps unwisely in the light of what happened later, they did not. Craig blamed genome researchers outside the NIH for standing in his way. ‘The extramural genome community did not want genome funding being used on intramural programs,’ he later told a congressional committee. Frustrated at the lack of support within his own institution, Craig looked for alternative sources of funds in the private sector. It was almost exactly the same time, ironically, that Bob and I were in discussion with Bourke. We chose one side of the line, and Craig chose the other.

Wallace Steinberg, chairman of the investment company

HealthCare Investment Corporation, funded Craig’s new institute to the tune of $70 million. TIGR was to be a non-profit-making operation, and Craig would be able to publish his work and so stay within the academic research community. There was a quid pro quo: in parallel Steinberg set up a commercial company, Human Genome Sciences (HGS), to market the discoveries of TIGR, and appointed William Haseltine as its chief executive. Haseltine had worked on the sequence of the AIDS virus HIV at the Dana-Farber Cancer Institute at Harvard University, and had also become rich through involvement in a number of biotech start-up companies. The deal was that Human Genome Sciences should have exclusive access to TIGR’s EST sequences for six months before publication, extendable to twelve months if sequences proved to be likely drug targets. Academic scientists would be able to look at the TIGR database freely after that, but the commercial company would have ‘reach through’ rights to any further commercial developments. Almost immediately the company sold an exclusive license for prior access to the information to the pharmaceutical giant SmithKline Beecham for $125 million. Craig, who had received shares in HGS from the outset, became a multi-millionaire almost overnight.

TIGR has remained a non-profit-making organization, and has been an important contributor to many publicly funded projects. Yet I saw the deal with Human Genome Sciences as compromising Craig’s academic integrity. I felt he wanted to have it both ways: to achieve recognition and acclaim from his peers for his scientific work, but also to accommodate the needs of his business partners for secrecy, and to enjoy the resulting profits. This apparent determination to have his cake and eat it set a pattern for what was to follow, when Craig launched a privately funded effort to sequence the entire human genome (see chapter 5).

Craig’s first priority at TIGR was to sequence human ESTs. Despite his frequent claims to have ‘developed’ the use of ESTs to find genes, he was by no means the first exponent of this strategy.

Sydney Brenner, who has never approved of large-scale genomic sequencing, had argued in favor of cDNA sequencing as an alternative to genomic sequencing during some of the earliest discussions about the HGP in the mid-1980s. (He used the excuse that ‘we should leave something for our successors to do.’) Sequencing cDNAs formed an important part of the work of his Molecular Genetics Unit in the late 1980s. What Craig did was to combine the EST strategy with the use of high-throughput methods based on the ABI fluorescence sequencers to speed up the discovery of candidate genes. This was no mean achievement, but was not a novel strategy.

TIGR did score a first, however, in sequencing the first free-living organism, the bacterium Haemophilus influenzae which causes chest and throat infections and meningitis in children. Craig was collaborating with Hamilton Smith of Johns Hopkins University, who had discovered restriction enzymes in the late 1960s and so launched the era of recombinant DNA technology. Smith and the TIGR scientists working on H. influenzae bypassed the mapping stage and shot-gunned the whole 1.8 megabase genome at once, relying on a computer search for overlaps to carry out the assembly, following the approach Fred Sanger and his colleagues had used for viruses. The TIGR scientists went on to use this whole-genome shotgun method to sequence other pathogenic bacteria such as Helicobacter pylori, which causes stomach ulcers.

Craig’s promotion of ESTs may not have found favor with the HGP, but it helped to launch a genome gold rush in the private sector. Suddenly the human genome looked like something you could sell. After HGS, one of the next on the bandwagon was the Palo Alto company Incyte Pharmaceuticals, later Incyte Genomics. After one of its co-founders, Randy Scott, read about Craig’s EST work in the New York Times, he switched the struggling company’s strategy from trying to isolate proteins that might make good drugs to producing a catalogue of human genes through EST sequencing.

Within a short time he had big drug companies queuing up to buy his product. Unlike Craig, Scott was a businessman first and foremost. The company would make its money first through selling access to the database, then from charging royalties on commercial developments from sequences that it had patented.

Five years before, genomic entrepreneurs had found it almost impossible to raise money. The Harvard scientist Walter Gilbert, who had shared the 1980 Nobel Prize with Fred Sanger for an alternative method of genome sequencing, had shocked his academic colleagues by trying to take the genome project private in 1987. But his idea of setting up a factory-like ‘Genome Corporation’ that would sell clones, data and sequencing services to academia and industry perished in the stock market crash of October that year. In complete contrast, from 1992 onwards genome scientists in universities found venture capitalists hammering on their doors. Many took the shilling. Gilbert bounced back by launching Myriad Genetics with Mark Skolnick of the University of Utah, dedicated to selling diagnostic and therapeutic products based on cancer-related genes. One of the genome project’s biggest grantholders, Eric Lander at the Whitehead Institute at the Massachusetts Institute of Technology, with Daniel Cohen of Généthon in Paris, co-founded Millennium Pharmaceuticals in Cambridge, Massachusetts. There were many others.

There is much potential for conflicts of interest as a result of these agreements. The natural inclination of a commercial company is to retain exclusive control of its product, whether through patents or through commercial secrecy. The goal of the genome project, as I saw it, was to provide as much information as possible that could be freely used by everyone, public and private, to advance our understanding and develop new treatments. I didn’t have a problem, and still don’t, with companies protecting their rights to the inventions they sell—drugs or diagnostic kits, for example—but I thought there was a real danger if they were able to gain exclusive rights to

the information contained in the sequence itself. That would mean that no one else further down the line would have any incentive to use that information in creative ways, and science and medicine would be poorer as a result.

From the earliest days of the Sanger Centre, we received visits from companies and entrepreneurs eager to make deals with us. My reply was always that we had nothing for sale, and for Bob it was the same. It worries me a bit that I may have missed opportunities for the U.K. by so doing. I was always aware that for Bob, as part of the vast capitalist empire that is the United States, this luxury could be afforded. There were plenty of other United States sequencing centers for the venture people to do deals with; but we were the only major one in the U.K., so what we did was decisive. The Wellcome Trust is a charity, so couldn’t trade directly, but we could have sold licenses to our intellectual property indirectly through its technology transfer section. Indeed, I was under some pressure at the beginning, from Bridget and Michael, to consider doing so. This led to an exchange with Michael one noisy evening in the Blackford bar at Cold Spring Harbor (why is it that one has to travel 3,000 miles to have meaningful exchanges with one’s colleagues?), when I said that if the data couldn’t be released freely then I would resign. As it turned out, my view was shared by many of the governors of the Trust, and so it became policy.

In the worm community these issues had seldom arisen. No fortunes were riding on worm genes, so everyone (more or less) was happy that information should be shared. The map, and increasingly the sequence, had increased everyone’s productivity and enlarged the field in a very effective way. But with the human genome it was a different story. The Sanger Centre began life in an environment in which commercial pressure was always going to be part of the picture. Those who were working to map particular human genes either expected to secure patents on them, or were terrified that someone else would beat them to it. It made for an atmosphere of

mutual suspicion. With our declared aim of mapping and then sequencing some large fraction of the human genome, it was inevitable that we were seen in some quarters as a threat rather than a resource.

An example of this was at the chromosome 22 workshop in 1994, when the Sanger Centre group led by Ian Dunham was accused by another mapping group of withholding data, and of using markers that had been discovered by others without giving appropriate credit. In fact there was nothing underhand about Ian’s approach—it was just that his perspective encompassed the whole chromosome rather than particular bits of it, and he had underestimated the desire of others to stake specific claims. The network of chromosome workshops had been set up by HUGO, primarily to collect and collate human genetic data; it was not easy for the participants to adapt to the new world of genomics. Personally, I saw the workshops as potentially valuable in integrating the two, but we had to get over the understandable resentment that we newfangled people were invading the geneticists’ patch. Many of the groups involved would have liked to move forward into mapping and small-scale sequencing. But that would take too much time and cost too much. It was already clear that increasing the scale of sequencing was going to be the key to getting the costs down. I hoped that we would find a way for everyone to be involved somehow, but we had to face economic realities.

A key step in gaining the trust of the community was to make it an absolute rule that all the sequence produced at the Sanger Centre, whether from worm, yeast or human, would be immediately released into the public domain. This was exactly what we had done previously with the worm map. Anyone who came asking us to sequence a particular clone had to accept that we would not keep it back to give them time to look for patenting opportunities: as soon as it was finished, it would be deposited in the public sequence databases, and out there for everyone to see. By doing this, we would not

only help science to progress; we would make the sequences in themselves unpatentable. Patenting opportunities would be open only to those who went on to do the real work of discovering what the sequences did and developing commercial products based on that understanding. Immediate release would also ensure that the sequence was straight away available to users, avoiding any wasteful duplication, and demonstrating that our own lab was not profiting from it—in the current jargon, it was being treated precompetitively. This had the practical benefit of both giving us immediate feedback on what we had done, and winning the support of the research community for a project that some had resented for taking so much money. In fact, sequencing projects account for less than 1 percent of the total biomedical research budget, and by making the data freely available we ensure that everyone can benefit from the results.

But all this, while truly a calculation that we made, makes the policy sound too purely pragmatic, or at least too short-term. The fact is that we’d come to realize that the genomic sequence we were producing and dealing with is more than a commodity. It is the essence of biological heritage, the instruction book for living things. We’d recognized it for the worm, spoken of it in our proposal for establishing the Sanger Centre, but now we had gone further. Now the information that was coming off our sequencing machines was the code of our own species. The only reasonable way of dealing with the human genome sequence is to say that it belongs to us all—it is the common heritage of humankind.