Endgame of Owning the Human Genome: Celera, Us (Score 1)

As Sean Eddy noted, Celera has a way to go before delivering a finished human genome sequence. This only slightly diminishes Celera's ability to widely acquire utility patents on the use of individual genes for developing therapeutics and diagnostics (drugs and tests for diseases), and the small changes (mainly single nucleotide polymorphisms) that are believed to differentiate individuals' tendancies for disease or amenability to a particular diagnostic therapy. On Celera's so-called 'speedy' results, to my recollection the original public genome project was purposefully designed to take a more thorough and slow approach to getting a high-quality product at the end. Closing the gaps and finishing the hard-to-sequence parts was expected to be scientifically important in the long run, but unglamorous and less interesting in the medium term. So rather than go for just the exciting quick results approach, the public project decided to get all the genes and the intervening, non-coding, DNA, and the regions of unknown purpose, thereby spreading the finising grind and the exciting work among all contributors. Although incomplete, Celera's data--combined with the public data--is quite well enough to start mining off the low-hanging fruit and wholesale patenting the human genome with their 80%return/20%of-the-effort approach. Moreover, this will allow Celera & co. to refine their workflow and tweak algorithms so that while the academmics are grinding out the higher quality data, so Celera's analysis pipeline will be the worlds most highly-tuned high-dollar pipeline for analyzing that work for patentable content. A good team of biologists motivated with stock-options at Celera, trained and seasoned on the public algorithms for genomic analysis, would be very, very hard to beat. By rumor, Celera has far more computational resources than the public projects, and I understand that they have been hiring aggressivley in the math and computational sciences, as well as luring cash-poor graduate-student biologists to help pore over their data. Furthermore, Celera has wisely put the drosophila (fruit fly) genome data in the public domain so that it can be analyzed and the data from the fruit fly will be available to figure out what all the human genes do by comparison (homology mapping from a known example is the #1 way to find out what a gene does. Fruit flies are ~>90% similar to humans at the gene level). Hence they have the computational horsepower to rip through their data with any and all available algorithms, and they have or may soon acquire the biological expertise to refine their approaches for gene recognition to a patentable level. It is an unfortunate point that the public domain groups can't restrict the licensing of their software for use by Celera. The core of genomic analysis software, from raw-dna assembly to what-does-it-do gene analysis, has been almost exclusively developed by the academics. And the balance of good computational genomicists remains in the public domain. Remember that the human genome data isn't worth a cold cup of coffee without the computational algorithms. To quicly sift through billions of dna base-pairs you must get in the ballpark for what any-random-gene in all-that-data does--with the computational algorithms. Only then can you hire a biologist to close in with a tight patent on what that gene might be good for.But the discussions I've had indicate that the academics have never made a serious fight on the basis of use of their algorithms (or data) to reign in Celera or similarly operating high-throughput data- and legal-mining compaines. Still, there are a few of us, like Sean E., who are working to put a bit of the human genome's intellecual property in the public domain. But I think it would be fair to say there is a 10x imbalance of human resources and a 100x imbalance of funding between computational scientists dedicated to putting the genome in the public doman and scientists employed/dedicated to seeing Celera succeed. Most computational biologists I know are bystanding, not keen on patenting the genome, also not interested in a 'white knight' approach to take on a juggernaut opponent, and mostly trying to crank out better tools for genomic analysis for use by anyone who cites their work. So the public computational efforts are likely triage for a limited portion of the genome--at best--if Celera's patents hold up. As I understand it (I'm no lawyer) all the wall-street payoff, all the intellectual property, comes from figuring out what the genes & features in the DNA do so you can file a utility patent. This law is not well-settled, though I expect Celera is working hard to couch their position in the single-gene-patent precedents which allow companies like Genentech (Human Growth Hormone, 15-20 yrs ago) and Geron (telomerase-like genes believed to control aging, mid-late 1990s). Celera's patents could be overruled or blocked by statute if it is deemed not in the public good. That may not be to the liking of large pharmaceutical companies who can out-bid others for the rights to the choice genes, though, so it would take a *lot* of political pressure to keep Celera from having rights to the human genome this way. In the days of dot.com dollars, Celera is huge to the scientists but tiny compared to the quiet interests (top 20 pharmaceutical firms) who would get a competitive advantage by keeping Celera's patent stable on retainer if they succeed. Still, you could email your congresspeople if this issue is important to you. Or maybe you can spare an aging $150K smith-waterman accelerator, 2.5 terrabyte RAID, and some high-wattage Oracle8i licenses? -RPO endnote: I hope Sean E. and others will correct any inaccuracies from my few thoughts in this tiny comment box.