Can you patent DNA? The Federal Court said yes on Friday -- at least in "isolated" form. Let me explain ...
In Cancer Voices Australia v Myriad Genetics
-- over the right to patent so-called isolated DNA that indicates a mutation that can cause breast cancer -- Justice Nicholas found in favour of patent owner Myriad.
It's no surprise -- it follows long-standing legal practice by the Australia Patent Office, and the legal situation in most other countries, including the United States where the most recent appeal on their equivalent to this case broadly agrees with the Australian decision (although that decision is currently being appealed to the US Supreme Court).
The Cancer Voices group argued nucleic acid, of which DNA and RNA are the most relevant examples, shouldn't be patentable as it occurs in nature -- the exact same form of DNA that is covered by the patent is found in the cells of women who are susceptible to getting breast cancer. Patentable inventions must be new. They also argued there is a public policy dimension to granting such patents, which they believe discourage research and may limit access to diagnosis and in discovering treatments.
Myriad, the owner of patent No. 686004
, argued the patent doesn't cover naturally occurring DNA but "isolated" DNA, for which considerable human intervention and effort is required to obtain and purify. It also argued the DNA sequence being patented is chemically different from natural DNA found in a person -- it's a shorter segment of what occurs naturally and therefore a different molecule -- but the judge wasn't convinced on that point.
The legal definition of patentability is clear: there must be an artificially created state of affairs having economic significance and can only granted to items that are new. The claims in the patent covered specific isolated segments of DNA and RNA. Think of DNA as the blueprints or engineering drawings, for making a robot perhaps; RNA is like the work orders that go out to order the specific parts for making the robot.
If a researcher finds a rare spider, say, which has an intricate joint in one of its legs that can be used as an exact model for constructing a robot arm, that exact joint that can be patented in metallic form for robot arm joints. Re-making the spider’s joint as a joint in a robot arm is patentable, despite it matching the function and arrangement of the natural joint. This is an invention because the step of discovering the natural version and then creating an analogue in metal is the artificial state required by patent law.
In contrast, a few weeks ago, a judge in another case in the Federal Court -- Research Affiliates v Commissioner of Patents
-- ruled that a patent for a stock market index (actually a method and computer program product for constructing "a non-capitalisation weighted portfolio of assets") was not patentable. The index was held by the court to merely consist of a set of data and therefore not patentable subject matter, and neither was a computer program to construct the index.
In this case, these methods were not be patentable because they did not produce "an artificial state of affairs" in the sense of any concrete, tangible, physical or observable effect resulted from operating the methods.
So while isolated natural products have been confirmed as being patentable, some business methods and software products may not be. Patent law is evolving along with technological developments.