Sibia Neurosciences, Inc. (“Sibia”) appeals the decision of the United States Patent and Trademark Office, Board of Patent Appeals and Interferences (“board”), In re Sibia Neurosciences, Inc., Nos. 90/005147 and 90/005326 (Jul. 30, 2004), affirming the examiner’s rejection of claims 15-52 of U.S. Pat. No. 5,401,629 (“the ’629 patent”) as obvious under 35 U.S.C. § 103. We affirm.
Obviousness is a legal determination based on underlying questions of fact. In re Gartside, 203 F.3d 1305, 1316 (Fed.Cir. 2000). We review the board’s conclusions of law de novo and affirm its findings of fact if they are supported by substantial evidence. Id. Findings of fact in an obviousness determination include, inter alia, the presence or absence of a motivation to combine references, id., and what a refer*315ence teaches, Para-Ordnance Mfg., Inc. v. SGS Imps. Int’l, Inc., 73 F.3d 1085, 1088 (Fed.Cir.1995).
Sibia argues that the board erred in finding that U.S. Pat. No. 5,071,773 (“Evans”); U.S. Pat. No. 5,747,336 (“Bonner”); and Sassine-Corsi, et. al., “Induction of proto-oncogene fos transcription through the adenylate cyclase pathway: characterization of a c-AMP responsive element,” Genes & Development, Vol. 2 (1988) (“Sassone-Corsi”), render claims 15-52 of the ’629 patent obvious. Those claims are directed toward a method for using a reporter gene to identify compounds that modulate the activity of G protein-coupled receptors (“GCPRs”). An increase in cyclic AMP (cAMP) levels is an intermediate step between the GCPR and reporter gene activation. Sibia argues that there is no motivation to combine the three references, that there is no reasonable expectation of success, and that the references do not teach the elements of the claim.
The board determined that the Bonner reference teaches a method for measuring ml muscarinic receptor activation by measuring changes in cAMP levels. The difference between the Bonner reference and the ’629 patent is that Bonner measured cAMP levels and not the reporter gene product. The reference states that “[t]he functional responses observed for one or more of the receptors include ... increases or decreases in cyclic AMP levels [which] could form the basis of a functional assay for the screening.” Bonner, Col. 5, lines 22-64. Thus, it cannot be said that the board’s finding as to what Bonner teaches was not supported by substantial evidence.
The board also determined that the Sassone-Corsi reference teaches treating cells to increase levels of cAMP and then monitoring the effects of the cAMP levels on promoters and reporter genes. The reference covers the steps in the ’629 patent between increases in cAMP levels and reporter gene activation. Therefore, the fact that the reference does not discuss GCPRs is irrelevant to its applicability to the step of the ’629 patent described.
The board further found that Evans shows the use of reporter genes to monitor receptor activation in general, contrary to Sibia’s contention that the reference only shows use of reporter genes to monitor activation of intracellular steroid hormone receptors. The board recognized the difference between GCPRs and steroid hormone receptors, but found that a person skilled in the art would not have considered the difference significant in this context. Thus, substantial evidence supports the finding that Evans provides a motivation to combine Bonner with Sassone-Corsi based on the idea that steroid receptor activation increases reporter gene activity.
Additionally, only a reasonable, not absolute, expectation of success is necessary to support an obviousness rejection. The board’s determination that there was a reasonable expectation of success in light of Bonner, Sassone-Corsi, and Evans, as well as the high level of skill in the art, was supported by substantial evidence.
Lastly, secondary considerations of non-obviousness such as long-felt need and commercial success do not overcome the examiner’s obviousness finding. Sibia failed to present evidence that its claimed method in the ’629 patent met a long-felt need. Further, Sibia’s declarations were directed toward the original claims 1-14, not the newly narrowed claims 15-52, and failed to provide the required nexus for nonobviousness based on commercial success. Therefore, substantial evidence supports the board’s finding.