Hacker News new | past | comments | ask | show | jobs | submit login
U.S. Says Genes Should Not Be Eligible for Patents (nytimes.com)
140 points by HardyLeung on Oct 30, 2010 | hide | past | web | favorite | 36 comments

Gene patents have not been a "problem" in the same way as software patents. And, unlike software patents, they are more likely to represent a significant R&D investment by the applicant.

(I am the author of the paper that showed how much of the human genome is patented. "Intellectual property landscape of the human genome", Science 2005. http://www.sciencemag.org/cgi/content/summary/310/5746/239 )

The brief filed by the government does not alter the law, it's just a letter to the court that says "this is the current, consensus opinion of the executive branch." The current law states that DNA (or any other naturally occurring chemical) cannot be patented unless it is significantly transformed from it's natural state. However, purification of the DNA is usually enough to render it patentable. That, of course, is not difficult with current molecular biology techniques.

But, like any other technology, patent applications covering DNA must meet many other requirements including non-obviousness and usefulness. Thus, the claims granted by the USPTO have become increasingly narrow as the field has matured and there is more prior art. That is, just like any new technology area the initially granted patents are broad, but they narrow over time.

So, that is a long way of saying, this might be a tempest in a teapot. Especially because most court watchers expect the Federal Circuit to reverse the District Court and keep the status quo.

Is non-obviousness actually applied in this sphere? In my corner of computer science, the non-obviousness bar is so low as to be basically non-existent. There are tons of patents where A and B each have significant prior art, but the completely straightforward combination "A+B" somehow gets patented.

It is applied, and it should also be in the CS space. But, as always, there are bad patents issued because they are examined by busy humans. In these cases, opponents of the patent have a number of options, most notably asking for a reexamination of the patent. During reexam, you can bring missed prior art to the USPTO's attention in an effort to invalidate the patent.

EDIT: I should have also mentioned, another option is to ask the courts to invalidate the patent (for obviousness, lack of enablement, or any other statutory requirement) if you are involved in a lawsuit. Even if you are not sued you can begin such a lawsuit by filing a "declaratory judgement" complaint against somebody who you reasonably feel might sue you.

As of 2007ish courts rely on the "Graham factors":

- the scope and content of the prior art; - the level of ordinary skill in the art; - the differences between the claimed invention and the prior art; and - objective evidence of nonobviousness.

e.g. http://en.wikipedia.org/wiki/Inventive_step_and_non-obviousn...

Gene patents have not been a "problem" in the same way as software patents.

I'd argue that the "problem" is that the entire idea that ownership, even temporary, can be claimed on a pre-existing string of bits (and not just some particular method of extracting or using that information) that exists inside all of our bodies, is downright offensive. Moreso because the techniques to isolate and exploit these bit strings have now become commonplace and all but trivial.

And, unlike software patents, they are more likely to represent a significant R&D investment by the applicant.

But the patent system is not a subsidy for R&D costs, it's supposed to protect innovative inventions. I'm appalled at the idea that a research group might be prevented from researching novel therapies that target a particular gene just because some other company happened to successfully isolate it using standard techniques ten years ago.

The claims are not just on the sequences, but can be on uses of the sequences, combinations, whatever. This is why genes are be patented in some cases up to 25+ times. Each patent owner claimed something unique that passed the USPTO's tests for patentability.

"However, purification of the DNA is usually enough to render it patentable."

Could you explain a little more about what this means? DNA in its natural state is impure?

Genomic DNA is always present with a ton of other biomolecules within a cell. Extracting pure DNA from the typical cellular mix of DNA, RNA, proteins, carbohydrates, lipids, etc. isn't much of a challenge anymore using modern molecular biology, but it is still necessary to study specific genes.

The DNA for a specific naturally occuring gene is only a tiny portion of the total DNA extracted from a cell.

Since 1990 or so, purification of this (relatively) small piece of DNA for a particular gene (from the enormous quantity of genomic DNA) has become pretty trivial in most cases. This is largely due to the invention of the Polymerase Chain Reaction [http://en.wikipedia.org/wiki/PCR] and the maturation of gene cloning techniques.

Do I understand correctly that just physically extracting a certain piece of DNA from a cell ("purifying" it) makes it patentable? That seems silly.

Do I understand correctly that just physically extracting a certain piece of DNA from a cell ("purifying" it) makes it patentable? That seems silly.

Yeah, it's ludicrous - imagine that I tried to file for a patent on using a random (but useful!) phrase in English on the basis that I was able to write a Perl regex to pull it out of a web page.

And I won't even qualify that with the usual "poor analogy" disclaimer - it's almost exactly what the current law allows for, just replace "random phrase in English" with "snippet of DNA", and "Perl regex" with "biochemical reaction".

Yes. That is basically the case.

I was under the impression that people were patenting genes as soon as they were identified and then going after anyone who did the research to discover what those genes were useful for because they found it first (even if they didn't know what it did).

There aren't many documented cases like you describe.

Not in the field of medicine. However, I wouldn't advise mentioning that gene patents aren't a problem to any farmers unless you want to hear a long spiel. The big agricultural companies go to great lengths to enforce their patents on certain strains of seed (including hiring PIs to spy on farmers, blacklisting farmers, and suing them for saving seeds).


DNA is just a sequence of bits, software is too.

Genetics has turned into a minefield because of patents.

Cross-licensing crap worse than even software.

It will probably take another decade if this ever changes, but I'm hopeful!

> DNA is just a sequence of bits, software is too

That's not the argument being made by DOJ. They're not saying that DNA sequences per se are unpatentable. On the contrary, they agree that man-made DNA sequences are indeed potentially patentable (assuming they're not "obvious" in view of other pre-existing DNA sequences, including for example genes in the wild). The brief argues only that if a DNA sequence occurs in nature, you shouldn't be able to get a patent on the isolated sequence.

That could be kind of tricky. If we want to make custom proteins, for example, the most straightforward way to do it is to build up a protein from bits of naturally-occurring proteins for which we already know the folded structure.

I'll give a more concrete example. Certain zinc finger proteins can bind to particular DNA sequences. By chaining together a few of these units, we can engineer proteins that bind to a particular sequence of DNA and promote or inhibit its transcription into RNA -- obviously a very useful thing to be able to do.

Do these custom zinc finger proteins count as patentable? On the one hand, we're mostly plagiarizing from nature. On the other hand, there are a lot of tough issues to solve in order to get to where we can reliably use these things. Personally, I think the bioengineering challenges are a bigger barrier to competition than a patent could be: if someone set up a company that could sell custom zinc finger arrays at a reasonable price, their expertise would matter a lot more than any patent.

That's where obviousness analysis comes in. It's tough to do in hindsight. That's why courts look to "objective evidence." For example, praise from other researchers, or extensive citation of the published paper, could be indirect evidence of nonobviousness.

(One of my partners at my former law firm told a story about representing a major university in seeking a patent for a biotech invention. It took a number of years for the PTO to do the initial examination of the patent application. The patent examiner perfunctorily rejected the application, asserting that the claimed invention would have been obvious in view of the prior art. My partner responded that in the years that the inventors had been waiting on the PTO to examine the patent application, they had been awarded the Nobel Prize for that very invention. Result: Obviousness rejection withdrawn. Of course, I could well be misremembering, but it nicely illustrates the principle.)

> DNA is just a sequence of bits, software is too.

I don't quite see how that plays into it. You can just as easily dismiss any physical creation as a collection of elements.

Although to be fair, every patented thing can be represented as a sequence of bits.

I disagree.

Sure, you can describe your patented process/machine/etc. with text, which you can in turn represent as a sequence of bits. But there's a difference between the patented thing being described with a series of bits and it actually being a series of bits.

Genes are purely a series of bits. But unless you want to specify the exact place where every atom goes through some crazy coordinates, you can't represent an actual machine as a series of bits--you can just describe it with a series of bits.

The government's actual position is more subtle than the NYT title suggests:

"While the government took the plaintiffs’ side on the issue of isolated DNA, it sided with Myriad on patentability of manipulated DNA."

How long is a piece of manipulated DNA? Would claiming prior art be mathematically feasible on all manipulated DNA?

The first question is similar to asking, "How long is a string?" Arbitrarily long.

I think that "prior art" and "obviousness" claims for certain classes of manipulations should (in a world run by biologists) be useful. For example, if you discover New Gene X and manipulate its DNA to tag its protein product with a fluorescent protein, that's pretty damn obvious because that's what we do with every other protein.

It's all obvious in the natural world (and living things reuse code without regard for law.)

I am, frankly, amazed that gene patents are now ineligible where software/business method patents remain so. This might be consistent with whatever theology the courts have sanctified, but it is the exact opposite of what makes sense for business and innovation.

Firms are forced to pay for their research toward FDA approval, and that approval is extended only to a very specific drug or treatment. The narrowness of this channel yields one of the only business situations where patent law, from an innovation perspective, makes sense.

Contrast this with business method patents and software patents: you can use a zillion different idea, each independently patentable, for any business or product, there is a much smaller barrier to entry, and there are many ways to accomplish the same thing. Patents go from being quite valuable for the innovator to being, at best, a distraction, and at worst, a minefield.


That's the first sane official statement about patents I've heard in a long time. There must be something I'm missing in this debate, because to me the fact that naturally occurring genes are not inventions is as obvious as that mathematical theorems are. (It's actually even more obvious, the genes can be plainly seen around us, when you know they're there.) It's as if software algorithms were written on the leaves of trees around us, but we needed to dye them to see the source.

Whether a synthetic gene is an invention is a different story to me. That's much more closely related to whether software is patentable. But no one would seriously argue that they could patent RSA if they turned a rock over and found the algorithm written down, no matter how many stones they had to turn to find it.

> It's as if software algorithms were written on the leaves of trees around us, but we needed to dye them to see the source.

Love the vivid analogy. But couldn't the same thing be said about any "method" invention?

Yes. It's not a black and white area. In principle, every method that works is just discovered, just like a theorem is.

To play devil's advocate - one could argue that allowing gene's to be patented would result in a vast amount of resources being invested, immediately, to decode and determine precisely what each of them do - with the rewards from such decoding being secured for 18 years (patent lifetime) - at which point anyone would be free to use the information.

In the absence of a patent incentive, the alternative, then, will hopefully be large government and other large public investments making up the difference.

I won't do the full analysis justice in a brief comment, but this is an empirical question that has some evidence.

The draft human reference genome was published 10 years ago. Prior to this, dozens of genes were discovered and patented by Celera. Celera gave free access to these genes to academics, but required payment for any for-profit activity. By one analysis, genes that were ever controlled by Celera, despite these relatively loose restrictions, as of 2009 had 30% less research done on them than those genes that were never controlled by Celera.


I worked for a very well known company who licensed some of the early genes that were patented by companies like Celera. It was such a painful experience and the arrangement caused tons of legal knots years later. When the license period ended I distinctly remember an email asking that in addition to surrendering all documents related to the genes, we had to "forget" anything we might have learned personally about them. By that time they were all publicly known, so it was a bit hard to tease out what was learned as a result of the patented genes vs. what you learned from a search at NCBI.

If you blog, I'd love to see that comment fleshed out in longer form. Your story sounds very interesting!

I don't have a blog because I have yet to convince myself other people would be interested in what I have to say. I might have to start one.

Ugh, anyone seen Food Inc.? It still doesn't solve the problem of Monsanto and their genetically modified crop, bullying small farmers.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact