Monday, August 17, 2009

Microsoft Patents Phylogenetic Comparative Methods. . . Say What?

I read my first (and hopefully last) patent application today, entitled Clustering Phylogenetic Variation Patterns. Although the title of the application sounds more like a late 1950s era manuscript on numerical taxonomy than it does a modern (Jan. 29, 2009) patent filing; it is nonetheless invoking considerable concern in the systematics community, as well it should.

According to this interesting article on the filing by Elizabeth Pennisi (published in last week’s Science), David Hillis first response to the news that Microsoft had filed a patent pertaining to evolutionary inference from phylogenetic trees was that it must be a joke akin to “The Onion article about Microsoft attempting to patent 1’s and 0’s.

The patent filing, by Stuart Ozer, claims invention of a variety of techniques already in wide use by systematists and evolutionary biologists – and (so far as I could tell) none of these inventions are original in quality. The whole patent filing can be read (at ones own risk) in its entirety here, however I have also chosen a few select passages for reproduction, below.

Among the claims of invention in this patent filing, the author purports to originate:

“a method of generating biomolecular clustering patterns”

“mapping at least a portion of the plurality of sequences to an evolutionary tree, the evolutionary tree including a plurality of nodes corresponding to the sequences in a hierarchical arrangement”

“counting evolutionary events in each of the identified plurality of positions at each identified node in the evolutionary tree”

“pruning the evolutionary tree”

“identifying the identified node as a leaf node based on the species count of the corresponding child nodes” for example if “the species count for each of the corresponding child nodes is less than a predetermined number”

“mapping evolutionary events”

“counting evolutionary events further includes: generating an event rate . . . wherein identifying related positions includes identifying related positions based on the event rate”

Without knowing the details of these “new” methods, the claims of invention are hard to evaluate. However, the language of the patent is so broad that it seems possible that if this ridiculous patent is ever approved, we might find that in doing what systematists have done for years, we will be infringing on a patent held by Microsoft Corporation!

Perhaps most nefarious about the filing is the following claim (pointed out by Roberto Keller commenting on the Myrmecos blog):

“Efficient and accurate identification of related groups of monomers of biomolecular sequences is important in achieving biotechnological advances in research and development. However, there is currently no efficient method for determining groupings of monomers in a biomolecular sequence, or among related interacting sequences.”

The author of the patent filing is apparently unaware of the last 50 or so years of research in this area (and should perhaps familiarize himself with a wonderfully interesting chapter on the subject in Felsenstein’s book).

For more on this story, see the following links:

Elizabeth Pennisi's article (link also above)
Systematic Biology
John Hawks weblog
Myrmecos blog
Kevin Zelnio's blog

9 comments:

Dan Warren said...

I read about this the other day, and it is indeed disturbing. Is there any organized effort to contest the patent application?

Leonardo de Oliveira Martins said...

The patent aplication with figures can be found here:
http://www.google.com/patents/about?id=gHezAAAAEBAJ&dq=clustering+phylogenetic

BTW, searching on google patents for "phylogenetics" returns several other disturbing applications. The language is so loose I cannot make sense out of them...

Susan Perkins said...

Ah, yes, one of these is held by one of my AMNH colleagues. See it here. I believe most will recognize this as "direct optimization."

Joe Felsenstein said...

My reading of the patent application is that what they intend to patent is the clustering of tip species (or internal nodes) on a tree using a vector of how many changes there have been above that point on the tree in each character. This a somewhat oddball thing to want to do.

My sources who are knowledgeable about Microsoft Research tell me that they are good to work for, allow you to work on what you want to, and don't constrain you from publishing. But they do insist that everything you do be patented (mostly to prevent Google from using it). If you make a tuna fish sandwich for lunch there, you patent it.

So I suspect that this patent does not mean that Microsoft is going to claim ownership of standard phylogeny methods.

Liam Revell said...

Leonardo - thanks for the link. I updated the original post.

Dr. Felsenstein - thanks for the insight!

Incidentally, after a note about this blog entry was posted to NESCent's twitter feed, Microsoft found it and began to follow our tweets - within 2 seconds!

Joe Felsenstein said...

OK, I have looked at Google Patents, using the search on "phylogenetics", and I am alarmed. There are a bunch of Microsoft patents, probably reflecting their insistence that their Microsoft Research people patent whatever they do (and not therefore necessarily indicating any large plans to use these patents to encroach on use of existing phylogenetic methods).

But a little defensiveness would not be a bad thing. Shouldn't the major systematics and molecular evolution societies look into making sure that standard "prior art" methods don't get patented?

Glor said...

It will be a sad day if every new method needs to be associated with a patent application.

Does coming up with an idea prior to a corporate minion who happens to file a patent provide protection against subsequent accusations of infringement, or does the person filing the patent get precedence? In any case, it does seem like our professional societies need to get on the case.

Joe Felsenstein said...

I am no legal expert but I believe the relevant term is "prior art": when you can convince the authorities (or a court) that the technique was in wide use by others already, even though they did not file a patent, you may be able to prevent a patent from being in force. See:

http://en.wikipedia.org/wiki/Prior_art

John Harshman said...

One major problem would seem to be in deciphering just what the patent claims are intended to mean. I'm sure they were made as vague as possible one purpose. But what, legally, would they mean? What methods would be encompassed by the claims? It's not at all clear to me, given the obscure language. And thus not clear to me whether we should or should not be alarmed.