Privatizing the Human Genome?

Principals behind joint-venture proposal and public effort seek to define relationships A private effort to sequence the human genome four years ahead of the Human Genome Project's 2005 goal could either compete directly with the federal project or meld seamlessly with it. Before any relationship between the two efforts becomes formalized, scientists and federal officials involved with the Human Genome Project must determine whether the private approach will work, who will own the data, how quickly that data will be released, and how complete a picture of the genome that data will paint. The private effort, a joint venture between Perkin-Elmer Inc., a Norwalk, Conn., manufacturer of sequencing equipment, and J. Craig Venter , president and director of The Institute for Genomic Research (TIGR) in Rockville, Md., would be embodied by a new company formed to sequence the entire human genome by 2001, and at an estimated cost of $150 million to $300 million. Faster, more automated sequencing machines recently unveiled by Perkin-Elmer, and a different sequencing strategy espoused by Venter make that target feasible. The federally sponsored Human Genome Project, which has a total estimated budget of $3 billion, is on schedule to finish sequencing the approximately 3 billion base pairs of DNA that make up the human genome by 2005. MIXED SIGNALS: NHGRI's Francis Collins has sent mixed signals about whether the public and private sequencing ventures can cooperate. Francis Collins , director of the National Human Genome Research Institute, cautions that the "cheaper, better, faster" image the private project has attracted (some might say courted) may be a bit of a mirage. He notes that the federal price tag of $3 billion includes the sequencing of other genomes, necessary to do the comparative work required to ascertain each gene's function. Also, the federal project's larger budget has paid for much of the research--including gene mapping--that the private venture will use as its foundation. Collins told The Scientist that only about $100 million has been spent on human genome sequencing thus far. He and others are unsure whether the private product will be better--only that it will be different, since the public and private efforts will use different sequencing strategies. That difference leads Collins to question whether the data from the two projects will be compatible. A test for whether the two sectors will indeed become "melded together," as Collins says, may be the sequencing of Drosophila, which the venture intends to complete as a test. Even the success of that project doesn't necessarily guarantee the success for the human approach, since the human genome dwarfs the fly's genome in size and complexity. However, failure--or the production of unusable data--from the private venture could doom further interaction. "If the strategy fails for the fly, it will almost certainly fail in the human." Finally, Collins and others concede that the method could be faster--if it works. However, cooperation between the as-yet-unnamed company and the federal project could speed the completion of the human genome even further, Collins and others agree. Venter reports that he discussed potential collaboration scenarios (see related story) with Collins, Harold Varmus , director of the National Institutes of Health, and others. However, whether that cooperation materializes remains to be seen. Reactions to the May 10 announcement about the commercial venture have been mixed. "There are significant uncertainties about what the product will be and what its availability will be," Elbert Branscomb, director of the Joint Genome Institute in Livermore, Calif., told The Scientist. "What those differences are and how significant they are are not that clear." Still, Branscomb suspects that, at minimum, the federally sponsored project will adapt to account for the private venture. The venture's potential impact--plans call for 230 automated sequencing machines to collectively churn out 100 million base pairs a day--is too potentially profound to ignore. "I think it would be surprising if the public effort did not adjust its strategy and investment." The scope and nature of that adjustment are yet to be defined. Theoretical scenarios range from a full-blown genome sequencing race, fueled by increased federal funding, to complete cooperation between sectors, perhaps with the private effort cranking out the rough sequence, while federally funded centers fill in the gaps and do comparative work to ascertain each gene's function. The extent of adjustment and cooperation may be worked out over the next few months. All sides may have until April 1999, when the venture is scheduled to kick into full gear, to define their roles. RAW RELEASE: Perkin-Elmer CEO Tony White intends to release raw genomic data from the project to the scientific community for free. Data release may be one of the biggest obstacles to a full-blown collaboration between the public and private effort. Michael Hunkapiller, head of Perkin-Elmer's applied biosystems division, admits that the instrument company will, by launching this new venture, also become an information company. The new company will sell access to databases containing information about genetic variations around single genes. Pharmaceutical firms are likely to be interested in such databases. Perhaps more alarming to potential public collaborators, the company plans to also sell patent rights of genes to pharmaceutical companies and biotech firms. Tony White, Perkin-Elmer CEO, says raw data will be released to the scientific community for free: "We plan to make the genomic information widely available." However, the frequency with which the company releases data to the public may be the key to the plan's acceptance by the scientific community. Branscomb notes that White intends a three-month lag between generating the data and releasing it for public consumption, although, "the international community has agreed to a daily release policy." That lag between discovery and release could be used to secure patents to genes before the information becomes public, Branscomb worries. "For the public effort, that is a significant stumbling block." J. Craig Venter, director, The Institute for Genetic Information Michael J. Morgan, program director for the Wellcome Trust of London, agrees. "We want data released every day that it comes out of the machine." The Wellcome Trust funds the Sanger Centre in London, which has produced about 30 percent of the data for the Human Genome Project thus far. Days after the announcement of the private venture, the foundation doubled its support for the Sanger Centre's efforts, although Morgan denies the higher commitment was a reaction to the new development. The press conference "did not leave me satisfied that this was an entirely benevolent, charitable act," Morgan says. "It is to make money." Morgan also doubts whether Venter's proposed method, the "whole-genome shotgun approach," will work for the human genome, because the approach has only been used on much simpler organisms and results in some gaps of information (See Hot Papers, page 11). That approach does not meet with the public project's goal. "Our objective is to end up with a reference gold standard representation of the genome--largely contiguous, with few if any gaps," Morgan explains. Branscomb, too, refers to the private project's intended outcome as a "rough draft" of the complete genome. Both note that an earlier Venter proposal to try this approach was denied NIH funding in 1997--although at the time, the new machines recently unveiled by Perkin-Elmer weren't available. Venter admits the method will result in some gaps, but emphasizes that the genome centers involved in the public project could fill those, if they so choose. William Haseltine, CEO of Human Genome Sciences Inc., a Rockville, Md., company that formerly provided funding for TIGR, agrees that expanding the technique's application from relatively small, simple bacterial genomes to the massive and complex human genome could prove difficult. The "shotgun" technique yielded gaps in the public database for some bacteria, so applying it to humans will result in more. "There are major scale-up issues." Branscomb concurs that the proposed approach has some formidable challenges to overcome. "In doing the shotgun approach, the difficulty in getting the little bits to assemble into a whole gets worse exponentially," as the size of the genome to which it is applied increases. He also notes that the human genome has more distributed and tandem repeats than bacterial genomes, which could contribute to more gaps. But the new team has assets that could help them overcome these difficulties--assets provided by the federally funded project. The information already in the federal database--especially back clones and end sequences--will provide the scaffolding, or reference points, to hang the bits of the human genome after it is fractionated into tiny bits. "That collection of data should help the process very much." In that respect, some collaboration already exists. Branscomb anticipates that, eventually, the federal and the private efforts will complement each other--in some as-yet-undefined way. The alternative is redundancy. "If we are truly redundant we should get out of the business and do something else." Editor's Note: The announcement that J. Craig Venter planned to lead an effort in sequencing the entire human genome--seemingly in competition with the federal effort--ignited a firestorm of reaction. Critics of the director of The Institute for Genetic Information (TIGR) said that the new effort would be incomplete, and that the corporate sponsors would hold the data hostage or sell it to the highest bidder. Immediately following the May 10 announcement, federal officials hinted in the popular press that the federal and private efforts would remain separate. As the news has sunk in, though, it appears that some adjustments in the federal program may be likely. Venter, during several telephone interviews with associate editor Paul Smaglik, sketches out some possible scenarios for collaboration. In the following edited transcripts of those conversations, Venter also defends his track record of data release, elaborating on how his experiences with Human Genome Sciences Inc, of Rockville, Md.--the company that provided initial funding for TIGR in 1992 and the company from which TIGR parted in 1997--shaped his decision to once again join forces with a corporate partner. Q: How do you see the relationship between this new project and the Human Genome Project? A: It remains to be determined. But let me tell you the strategy we've followed. As soon as we decided to announce this, we briefed Dr. Varmus, Dr. Collins, Ari Patrinos, Martha Krebs and the DOE and indicated that we were trying to announce this in a way that is not attacking the work that's been done thus far. In fact, we want to build on some of that work. We would love to do this in a scientific, collaborative fashion. There's a lot of ways that make sense. People are worried about us being able to close the genome. Our strategy is going to leave little microgaps everywhere. We're delighted to work with the existing genome centers to build the scaffolding faster, have it go together faster, and have the gaps closed, if [the gaps are deemed] important to close. We're looking for interactions and we're looking for collaborations. It's not clear that that's the way it'll end up going just because this is being greeted with a little bit more hostility by some of the genome centers than we thought might happen, but we're still open to it. Q: How do you account for some of that hostility? Do you think people feel a little threatened by this project at this stage? A: Any time there's a new fundamental paradigm change, it certainly impacts people until there's an adjustment for it. The other point that we're arguing with genomics is [that] getting the first referenced human genome is the beginning. It's not the end of anything. Part of the human genome project from the beginning was to do the mouse genome to overlay it on top of the human to try and understand it. If all the same genome labs switched immediately to doing the mouse genome, the worldwide understanding of the human genome and of biology would move forward by a decade. So, it's not like there's not things to sequence or do, even if people will give us credit that we're going to do this and can do it. If they don't, they can keep spending money and doing what they're doing. Obviously, if I were on the other side of this, I would feel initially upset and threatened, too. I think it's an easy emotion to understand ... . We're not trying to do this as a power shift or anything else. Obviously, that's how some people are interpreting this. Q: How do you account for persistence of concern about how widely accessible and available this information will be? A: It persists because some people are trying to make that an issue as a sort of a straw man to attack. You can attack someone's integrity. And how do you prove you're really honest? I've released more genetic information into the public databases than any other scientist in history. I don't think I need to prove that I'm going to release data into the public domain. Q: How did your relationship with HGS color this decision to once again take on a big corporate partnership? A: It colored it tremendously. I learned from my mistakes. I didn't have to convince any of the senior management at Perkin-Elmer that we had to release the data for free to the scientific community. It's a perceived issue in the community. It's not an issue at all in terms of what we are doing. The people at Perkin-Elmer have just been phenomenal. The ethic from the beginning was, we're going to sequence the entire human genome. We have an obligation to make that data available for free to the world community. The business plan is not based on holding the data hostage or keeping it secret like some of these other genome companies. That's what's sort of hard for people to understand--if you give the data away for free, how do you make any money off of it? We're going to release the sequence data that we are generating. We are going to release the contigs as we are going. And in the end we are going to publish a complete human genome sequence. I can understand why people would be skeptical about data release. Because other companies have huge troves of EST sequences from the previous technique that I developed and they keep that data secret. Scientists can't have access to it. It was definitely not my choice. Q: How much of a coordinated effort is there going to be between the government-sponsored effort and your effort? A: I'm obviously not the one who can answer that question. I can tell you what would be ideal for me--a collaborative effort to get it done even faster. What we'll have is the ability to do the high throughput sequencing and assembly very quickly. It would go even faster with the assembly if there were more genetic markers on the map, more back-end sequences contributed and novel strategies to deal with some of the regions that are going to be very refractory and difficult like telomeres and centromeres that were basically being ignored by existing genome projects anyway. The most important thing would be to accelerate doing the mouse genome to help make the genome sequence interpretable--that's what's going to really define all the regulatory regions. And that's part of the Human Genome Project anyway. What all this together would do is greatly accelerate the advances across the board. We're open to working with them on any level like that. I can't do anything but be available and willing to do that. Q: What would be the ideal scenario for cooperation between this venture and the existing Human Genome Project? A: There's any number of scenarios. But certainly anything that fundamentally moves the overall picture of genomics forward. Just getting the human genome sequence is only one phase of it. If you look in GenBank, there are only between 5,000 and 6,000 full-length cDNA clones. So, we're at least 55,000 short of the genome. It's those kinds of projects that are best [done] in a distributed university lab environment. It's not easy to set up one big genome factory to do those. There are 55,000 individual projects to do. There's lots of things that make use of the different types of infrastructures, that would enhance what everyone can do. Q: Is there going to be some concerted effort to work out relationships with other labs? A: I think in part there will be because we clearly let it be known that we are interested in having such positive cooperation and collaboration. We've been approached by the heads of some genome labs who would like to work with us on specific chromosomes. We're going to continue to have dialogues and participate in meetings reaffirming our interest in collaboration and good interactions with people. Some people are a little hot under the collar now, but I think they'll get over it in time.

Privatizing the Human Genome?

Interested in reading more?

Become a Member of