CCL: patents in computational chemistry and cheminformatics

 Sent to CCL by: "Mike  Gilson" [gilson .. verachem.com]
 Since this discussion begin with a discussion of a patent application with my
 name on it, here are my 2-3 cents!
 > Anyone who has experience with software pattents in cheminformatics,
 > computational chemistry and the likes?
 Patents are actually quite common in this area: go to the US Patent Office
 search site (http://patft.uspto.gov/netahtml/search-bool.html)and search
 for "accelrys", "tripos", "generalized born",
 "protein-ligand docking", for example. You'll find patents from both
 companies and universities.
 > If there were a repository for software that is in a public place, like
 > a scientific journal (e.g. JCC or JCTC), where one could upload new
 > versions of the code (together with a one-page description of the software)
 > then there would be a public track-record of software that can be
 > used in patent-cases as proof of prior art.
 Patenting the obvious or the known is definitely a problem in the US. The patent
 examiners seem to check for prior art only in their databases of prior patents.
 They don't look at the literature or the marketplace. Unfortunately, they
 probably also would not look in a software repository.
 > There is prior art for the method described, for example,the Qeq method of
 Goddard and Rappe.
 Like pretty much every other patent, this one builds on the work of others, and
 is patentable only insofar as it goes beyond what has been done in the past.
 Best regards,
 Mike