Bag8ger, gunnabe, Virgil...A commendable effort, however, if frog were really interested in the truth of the matter, he/she would have read the Patent Application by now. Clearly, he/she hasn't.
I also fully recognize that he/she is either unwilling or unable to understand/acknowledge the text that follows, but for the benefit of the intelligent and/or interested readers I offer the following FROM THE PATENT APPLICATION:
...the results demonstrated that a Lipitor® patient could be classified into the non-adverse response (muscle reaction) group with 96% accuracy, and that 98% of the cases (individuals who exhibited muscle reactions) were properly classified.
The statistical conclusion from the above text is that errors in the classifer fall on the side of placing some non-adverse responders into the adverse responder group. This would certainly not be unexpected since the vast majority of statin users, 95%, are non-adverse responders to begin with. But by the same token, fully 98% of the cases, those individuals who exhibited muscle reactions, were properly classified.
Using the example you have been discussing, this means that while a small number of individuals will incorrectly be told they are at risk of developing myalgia and be placed into the adverse responder class, only 1 of the fifty patients that IS predisposed to develop myalgia would have been placed in the non-adverse response group. In other words, by using the classifier to "identify" the adverse responders, the rate of myalgia incidence could be reduced from 50 in 1000 to 1 in 1000.
Since the goal of a classifier is to prevent adverse reactions, I am hard pressed to understand how this classifier would not be useful to a physician trying to protect his patients.
And as far as the "straw man" frog is attempting to establish by making a company PR the definitive, end-all piece of information (and mistating the contents on top of it) versus either the Patent Application itself OR the JPG Article, I would simply point out that the information in the PR is, necessarily, a "subset" of the information contained in the Patent Application and Article, not the other way around.
frog...try to understand the difference between a public relations piece, a patent application, and a scientific study and resulting paper. Quite simply, out of the three, the PR will always contain the least information. Why is that so difficult for you to comprehend?
As always, this is MHO. Do your own DD and make your own investment decisions.
Later,
W2P
AI
