Listeners to BBC Radio 4's "Moral Maze' recently heard the well-known geneticist, Prof Steve Jones, say that the dilemma faced by the scientist was the difficulty in distinguishing when one knew almost everything and when one knew almost nothing. Jones went on to say that, in his view, in the case of biology we know almost nothing. Add in the corporatization of the bio-sciences and...
The Great Gene Debate
By Matthew Herper
Wednesday February 21, 4:07 pm Eastern Time
First the presidential election needed to be recounted, now the very book of life. A former colleague and current rival of Celera Genomics President Craig Venter wants a recount.
William Haseltine, CEO of top biotech Human Genome Sciences (Nasdaq: HGSI - news), insists that both Celera (NYSE: CRA - news) and the international Human Genome Project counted far too few genes in their maps of the human genome.
When Celera and the international Human Genome Project unveiled their separate drafts of the human genome on Feb. 12, one of the biggest surprises was the number of genes. Previously, many scientists believed there were 100,000 genes. Now, it looks likely that there are a third as many. Celera says there may be as few as 26,000 genes.
Human Genome Sciences (HGS) and another genomics companies have long advertised that they had found more genes than previously announced--a lot more. The Rockville Md.-based firm says it has discovered 90,000 genes, and Palo Alto, Calif.-based Incyte (Nasdaq: INCY - news) long claimed some of its databases contained 140,000 gene transcripts.
Now HGS' chief is as mad about this alleged miscount as Al Gore was about the Florida debacle. ``This is not a trivial exercise,'' Haseltine says. ``It's vital to the future of medicine. A claim that substantially cuts the number of genes is a serious mistake. It is a mistake that, if people act upon it, can slow medical progress.''
Any impediment to medicine also slows the growth of emerging drug companies like HGS and companies that sell information to drug companies, like Celera and Incyte. A lot is riding on the genome. If Haseltine's right, a major advance is flawed. This will have ripple effects throughout the sector. If he's wrong, he and his company both look bad. The number of genes needed to make a human being is a matter of fact. Somebody must be wrong.
Incyte has a less-than-earth-shattering explanation of the differing counts: Comparing its database to Celera's is comparing apples and oranges, says Incyte CEO Roy Whitfield. Incyte, which counts every major drug company but Merck (NYSE: MRK - news) among its subscribers, doesn't collect genes as they are written in DNA. Instead, it collects the way genes are written in a similar chemical called RNA, which carries the information in the gene to the part of the cell that makes proteins. Each gene, as written in the genome, can probably make several different proteins.
HGS used a similar technique to find many of its gene transcripts. These gene sequences are called expression sequence transcripts, or ESTs. But Haseltine is not willing to play semantic games with the word ``gene.'' To Haseltine, comparing HGS' genes to those counted by Celera and the Human Genome Project is comparing ``apples to apples.'' But only 60% of the apples in HGS' databases are in the sequences published earlier this week. ``They miss a lot of genes we know make proteins,'' Haseltine says.
Based on HGS' databases, Haseltine says there must be between 90,000 and 120,000 genes.
Could this be true? Celera's Venter, who previously estimated that there were 50,000 to 80,000 genes, says HGS and Incyte were ``ignoring fundamental aspects of science.'' Then, he argues, they became involved in a race where the company with the most genes won. If they were blind to science, then they were free to find hundreds of thousands of genes.
Gene sequences found by methods other than sequencing the whole genome easily are either all or part of genes, Venter argues. The genome is a cluttered desk, where most of the clutter is not protein-coding genes. The body knows where to find the genes; researchers don't. The argument here is whether 30,000 genes make up 1% of the genome, or 90,000 genes make up 3% of the genome.
The lower number is not written in stone; Celera estimated a range with 40,000 at its upper end. Venter says it's always been an ``important caveat'' that the number could move back up. Like Haseltine, Kari Stefansson, CEO of Iceland's deCODE Genetics , doesn't believe that the programs used for gene counting could have possibly found every gene. As Scott Morrison, national director of Ernst & Young's life sciences practice, puts it, ``I think that number will move around a lot, but it won't go back to 90,000.''