Thanks to the Patnews mailing list, I was made aware of an alarming patent application from Genentech (WO2004060270, for intellectual property aficianados following along at home.) It's pretty unremarkable, if you just look at the front page. Two inventors, and a bland one-sentence abstract: "The present invention is directed to compositions of matter useful for the diagnosis and treatment of tumor in mammals and to methods of using those compositions of matter for the same."
Generic as can be - that sentence would do, if you were pressed for time, for any of several hundred patent applications that come out every year. Walking around the patent landscape, you'd never stop to pick that one up. But just try it. . .hmmm, seems to be buried in the ground a bit. Dig around, and it's larger than you think. . .you're going to need a shovel to get this thing out. As you go on, this nondescript little pebble turns out to be a bump on the anchor of an entire buried aircraft carrier, and a fleet of diesel-powered backhoes aren't going to be enough to unearth it.
This is just how you feel if you're foolish enough to try to download this thing. My company subscribes to a service called Micropatent, which for a fee gives you access to just about every application and issued patent you could want. Check-mark this one and they cheerfully inform you that it's five thousand and fifty-four pages long. Better make sure that the printer is still under a service contract.
It turns out that this behemoth contains over four thousand gene sequences, presumably the compositions of matter that they're talking about. Now, this would have (almost) made sense eight or ten years ago, in the land-rush era of genomic sequencing. People were flooding patent offices with all kinds of flimsy applications as soon as they got an open reading frame. But a few years ago, the various authorities began to tighten up on this sort of thing.
They wanted to see a lot more enablement of the claims that were coming in - no longer could you say: "I claim gene sequence ATCGAAGTA, etc., and all the diagnostic methods for whatever diseases turn out to be associated with it and any treatments that make someday make use of it and whatever else might end up turning a buck, that too." But that looks like just what Genentech is saying here. The patent has a priority date of October 2002, long after this kind of nonsense was supposed to have packed up.
What's going on? It's no afternoon's work, putting together a five-thousand-page application, and prosecuting it is going to get pretty expensive for a practical joke. I have to presume that they have a reason, but it's hard to see what it might be. The only thing I can think of is that they want to spray down all of these sequences as prior art.
Perhaps Genentech has an idea that there are some useful things in this haystack, but no way to be sure which ones they are. They know, appearances aside, that it's not 1994 any more, and they won't be able to get real patent rights for such a fishing expedition. So if they can't own them, well, here's the way to make sure that no one else can, either. Publish 'em openly, and they're non-patentable. The only problem is that no journal would accept a paper with a title like "Forty-Five Hundred Gene Sequences That Might Be Of Use Someday", with five thousand pages of supplementary material. Heck, they wouldn't have even taken that in 1994.
But the patent office will take it. They'll take most anything if you pay the filing fee, and the application will publish in a year or two no matter how hideous it is. It'll be interesting to watch the prosecution of this one. It's hard to imagine that Genentech is seriously going to take thing thing on and let it nationalize. Nope, my bet is that they're just going to drop it. The mission has already been accomplished.