On propellers, rivers, Albert and USPTO

Last Sunday's New York Times Business section had an inspiring article on Jay Harman, an inventor whose time has finally come. One paragraph is a good example of this enjoyable article. After informally studying vortexes for several decades, he went to his bathtub and, with a bit of cleverness, was able to create a cast from the vortex generated by water flowing down the drain. Then he used the cast to help redesign the rotating parts, or impellers, used in pumps and other devices to move fluids. Since then he has applied his approach more broadly to redesign all kinds of devices that move fluids and gases. Einstein was always fascinated by the laws of fluid movement, especially the flow of a river. On the contemporary note: when you see vortices of water flowing down the drain, you can imagine all the money wasted by this administration on fiddling with US Patent and Trademark Office.

Microsoft spends big on a patent typo

Microsoft has produced this fine example of how to develop patent claims: In November 2006, Microsoft filed a patent application of dubious merit, with the following first claim: Patent Application No. 20080134338 1. A method comprising:mapping a visual invariant to a program invariant; and discovering inputs to GUI logic that includes a user actions and a execution context to cause the program invariant to be violated. In June 2007, after millions of dollars of additional research, Microsoft filed a patent application (of dubious merit) which is a followup after a lengthy negotiations with at least 2 patent examiners having the following claim worded subtly different: Patent Application No. 20080133976 1. A method comprising:mapping a visual invariant to a program invariant; and discovering inputs to GUI logic that includes a user actions and an execution context to cause the program invariant to be violated. Having the indefinite article a may imply (as an example of one of the possibilities) the existence of context, which is not new, nor obvious, nor useful; as opposed to the an article implying the intended claim to a type of execution context, which may let the application to have (some) value. This is the millions of R&D dollars that Microsoft is spending to fix the rest of the grammatical errors in a claim, while choking patent examination pipelines with this crap.

Some belated news on licensing (non-computer) mice

There must be a reasonably robust market in pharma to license in animal models of diseases. There's really only one licensing model that works - the annual fee.

I have structured these agreements two ways. The first is a license to a *breeding pair* of mice. This pretty much grants a license to use the animals at whatever rate/volume necessary for a representative period of time. Start with three years. The upfront fee can be sizeable-I saw one that was US$150k, if I remember correctly-with future annuals much lower. The median license fee was probably about US$50k. Keep in mind that the fees are directly proportional to how useful (duh!) the model is. Yet another model for cancer will likely be less valuable as a quasi-validated model for obesity. Models that have not been published upon are much less valuable than those that are. The other way I've licensed mouse models is to license a finite number of mice. This requires a lot of coordination with the PI and their willingness to house/breed the mice and ship them to the end-user when requested. It's generally quite a pain, but it provided an incredible amount of control over the resource. The fees are lower (think US$20k for 10 mice), but there's always a chance to "reorder". Just make sure that you send all of one gender, as teaching abstinence to mice is kind of tough. ;-) Regarding patenting - DON'T. Think about the financials. You can generate a nice income off of the mice without having an issued patent. You also want to license the model non-exclusively to maximize its use, with the side benefit of all that extra revenue. Plus, you effectively can control much of the early access to the technology. I don't see any advantages to patenting, unless you want to add a count to your internal metrics. Besides, by the time a patent issues it's not unlikely that a better/newer model will be created that people want to try. A couple of final thoughts. In the license, make sure that you are licensing for the specific/published use. I would suggest taking the "easy" approach and not even attempting to get royalties. It's very unlikely that you'll get them and you might as well speed up the licensing process. Request that if additional applications are discovered that they are reported back to you. Ensure that the mice are NOT allowed to be cross-bred to seek out additional traits, unless you have joint ownership in the resulting animal and full ownership in terms of distribution (pay them a percentage of the take from new models). Oh, and be sure you understand the provenance of how the model was created (i.e. it wasn't a result of cross-breeding another institutions' mouse with one in your PI's lab).