Appeal No. 1996-2240
Application No. 08/105,482
markers of controls and individuals with the disease.” See page 15 of the Examiner’s
Answer.
Because fingerprint patterns and metabolic profiles are distinct properties or
features with no readily apparent connection, we infer that the only nexus between
Matson 1987 and Seltzer is that both references are concerned with the classification of
neurological disorders. We cannot agree that this alone provides the requisite reason or
2
suggestion to combine the references in the manner proposed by the examiner. A bare
assertion that it would have been obvious to analyze any biological sample or parameter
using any statistical model previously used to identify the presence of a neurological
disorder is insufficient. Further, appellant’s disclosure teaches that statistical models are
not interchangeable in the claimed method. The examiner has not explained why
frequency distribution probability analysis would have been selected over other models,
such as linear regression analysis, stepwise regression analysis, or cluster analysis, which
cannot successfully distinguish between disease and non-disease populations. See
pages 19 through 21 of the Specification.
Again, we find no reason stemming from the prior art which would have led a
person having ordinary skill to the claimed method. In our judgment, the only reason or
2As stated in Pro-Mold & Tool Co. v. Great Lakes Plastics, Inc., 75 F.3d 1568, 1573, 37 USPQ2d
1626, 1629 (Fed. Cir. 1996) (citation omitted), “It is well-established that before a conclusion of obviousness
may be made based on a combination of references, there must have been a reason, suggestion, or
motivation to lead an inventor to combine those references.”
13
Page: Previous 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Next
Last modified: November 3, 2007