June 23, 1998
It's a Three-Legged Race to Decipher the Human Genome
The Race to Decipher the Human Genome
By NICHOLAS WADE
o one would ever invite the human genome to a party. It is long winded, highly repetitive and frustratingly full of intriguing stories that it has no idea how to tell in an organized way.
Still, the genome has suddenly become hot property. There are new ways of making it talk, and it holds a lot of secrets. Some are merely worth billions to drug companies. Others are priceless, such as those that tell how and when the human species evolved.
One plan is to do a 'first edition' and then fill in the gaps.
The effort to decipher the 3 billion letters of DNA that are the human genome unexpectedly became a race last month when the Federal Human Genome Project acquired a rival, a joint venture of the instrument maker Perkin-Elmer of Norwalk, Conn., and Dr. J. Craig Venter of the Institute for Genomic Research of Rockville, Md.
Venter, a competitive scientist who is not averse to tweaking the establishment's tail on occasion, announced on May 10 that he would start and essentially complete the human genome in three years, four years ahead of the government's target date of 2005. Because of his previous record in deciphering the small genomes of bacteria, his claim had substance, although some academic scientists say his method will not work and others that it will not be sufficiently accurate.
The response to the Perkin-Elmer-TIGR threat by those involved in the Human Genome Project is still in a state of development. The choices: Get the full genome done even faster, or do a quick, "first edition" version to pre-empt the Venter end run.
On the public front, Venter has been persuaded to play down assertions of finishing first and join in professions of collaboration. He told the House subcommittee on energy and environment last week that the Human Genome Project's sequencing effort was not a waste of money (as it might seem if Venter should finish first) and that "by working closely together" the two efforts would "complete the sequence sooner than anyone ever imagined."
Dr. Francis Collins, director of the National Institutes of Health's part of the Human Genome Project, told the House subcommittee that "the private and public genome sequencing efforts should not be seen as engaged in a race."
In practice, however, it is difficult for the academic scientists financed by Collins' office to ignore Venter's challenge. One response is the suggestion that the university sequencing centers should produce a quick first edition of the genome on a time scale competitive with Venter's. This means, in essence, doing all the easy parts of the genome first and then filling in the gaps.
One advocate of this approach is Dr. Leroy Hood, a biologist at the University of Washington at Seattle and a co-inventor of the DNA sequencing machines sold by Perkin-Elmer. "Low pass sequencing would allow us at lower cost to march much more rapidly through the human genome," Hood said in an interview, referring to a strategy of using fewer overlapping sequences than usual.
The strategy would also be less accurate, but its error rate would be similar to the natural rate of variation in DNA among individuals and not of significance in defining the genes, Hood believes.
If further tests showed that a first edition of the genome was feasible, "We could go to Congress and say we could get something useful done very quickly," Hood said.
Before Venter's announcement of his unilateral effort, academic scientists had tended to rebuff suggestions for quick solutions, favoring high quality and accuracy over speed. Always at the back of their minds has been the fear that if the obviously interesting parts of the genome were sequenced first, money might evaporate before the remaining regions could be completed.
Only 3 percent of the DNA in the human genome codes for genes, and much of the rest is apparently meaningless clutter that evolution has neglected to clear out of the genetic attic. Biologists want to sequence the noncoding DNA because it may well contain unexpected information, and in any case its variations help tell the story of human evolution.
Collins noted in an interview that it might be worth trying to focus for now on the gene-rich regions of the human chromosomes, leaving the gene-poor regions for later.
Though many scientists now favor accelerating the genome project, some assert that Venter's effort is a publicity machine that will not deliver promised results. Warning of the danger of overreacting, Dr. Maynard Olson of the University of Washington at Seattle told the House hearing last week that the Perkin-Elmer/TIGR project would produce data of poor quality, inadequate for the long-term needs of scientists.
The Venter strategy "will encounter catastrophic problems" when he tries to assemble component parts into a full sequence, resulting in more than 100,000 serious gaps, Olson told the House subcommittee.
Venter is not the only player whose actions need to be taken into account. Another participant is the Sanger Center in Cambridge, England. The center is financed by the Wellcome Trust of London to sequence one-third of the human genome in collaboration with the Federal Human Genome Project.
Its director, Dr. John Sulston, is committed to producing a full, high-quality archival sequence. Dr. Malcolm Morgan, the trust's program director, has said it would finance the Sanger Center to do the whole genome if its American partners should recede from the goal of a complete, high-quality sequence.
Players of another sort are companies like Incyte and Human Genome Sciences that have specialized in fishing out and sequencing just the segments of DNA that code for genes. The method for stripping out the genes was developed by Venter when his nonprofit institute was allied to Human Genome Sciences.
Dr. William Haseltine, president of Human Genome Sciences (also in Rockville, Md.), says that his company has already pulled out and sequenced more than 90 percent of the human genes. He has infuriated many university scientists by saying he sees no practical reason for sequencing the DNA in between the genes.
He also disputes the position of the Human Genome Project's scientists that their finished DNA sequences have fewer than one error in 10,000 base pairs of DNA; comparing them with his own gene sequences, Haseltine says that the sequences deposited by university scientists in public databases have one error in every 500 bases.
Haseltine's statements cannot be verified because Human Genome Sciences does not publish its data. The principle that motivates the Human Genome Project's university sequencing centers is the antithesis of Haseltine's. The centers' scientists believe that the human genome data should be in the public domain. The centers make no private use of their own data, and by agreement publish it immediately for the free benefit of other researchers.
Haseltine's claim that he has already found most of the human genes is strongly disputed by his former partner. Venter said that Haseltine's collection probably consists mostly of the common house-keeping genes that are easy to fish out because they are so active. Genes that flick on less often, such as interesting regulatory genes, are less likely to be picked up, Venter said. Haseltine said his method picked up rare genes too.
Some deride the Venter effort as a publicity machine.
Neither scientist is without self-interest. Haseltine's company would benefit if all rival sequencers vanished; Venter has presumably convinced his sponsor, Perkin-Elmer, that at least some patentable genes remain to be found so as to help defray the company's investment.
As if to emphasize their continuing differences, as well as how little still is known about the human genome, Venter and Haseltine differ by a factor of almost two as to how many genes it contains. Venter estimates there are 60,000 to 80,000 human genes; Haseltine insists there are 100,000 to 120,000.
Other scientists say the dispute demonstrates another reason for sequencing the whole genome.
The dilemma for those involved in the Human Genome Project is that they need to produce a first edition to protect themselves from Venter, yet such an effort would probably delay the goal of obtaining a complete sequence by 2005.
Though the university sequencing centers have made big strides, it is far from certain that the 2005 goal will be achieved. None of the centers have achieved the production rates promised two years ago, and some are much less efficient than others, according to production and cost figures reported last month by Elizabeth Pennisi in Science.
The academic sequencers' problem would be resolved if Congress provided more money so they could pursue fast and slow approaches in parallel. Some scientists who at first resented what they saw as Venter's declaration of war have now come around to believing his work might help jump-start their own efforts. But Rep. John Edward Porter, R-Ill., chairman of the House Appropriations subcommittee that approves the National Institutes of Health's budget, said in an interview, "I don't see it as a race, and so I don't see any need to change our funding."