To the Editor:

Some DNA sequence patent holders may be feeling like Mark Twain when he read his premature obituary. We believe the patent article by Miles Yamanaka1 in the October issue entitled “A nail in the coffin of DNA sequence patents?” is unduly alarmist. The headline and final sentence both imply that the decision by the Board of Patent Appeals and Interferences (BPAI) on the patent application of Kubin and Goodwin (application no. 09/667,859) threatens all DNA sequence patents. This is misleading because it is overly broad. In Kubin, the BPAI does suggest a higher standard for nonobviousness2, a criterion that the US Court of Appeals for the Federal Circuit unwisely rendered largely inoperable for DNA sequence patents in its 1995 Deuel decision3. In the eyes of most analysts, Kubin is a sensible corrective. But, even assuming that the Federal Circuit goes along with the BPAI's reasoning, precisely how Kubin will affect DNA patents as a whole is hardly clear.

Read narrowly, the BPAI decision precludes only claims on DNA sequence based on prior characterization of a protein's amino acid sequence. On that reading, Kubin merely captures a judgment that deriving a nucleic acid sequence from a corresponding amino acid sequence is straightforward to those with ordinary skill in the art, despite some degeneracy of the genetic code. (Yamanaka acknowledges the possibility of this narrow reading when he states “the Kubin decision will make it harder to obtain claims to a polynucleotide encoding a protein when that encoded protein is already known” [emphasis added].) But claims to DNA sequence derived from amino acid sequence are mainly confined to some 'first generation' gene patents based on cloning genes for known proteins. Most DNA sequence patents that we study in our work, for example, are not based on prior characterization of a protein, but start from a genetic discovery or DNA sequence variation.

Even if Kubin is read more broadly, to render invalid all composition of matter claims to DNA sequence patents where the procedure for finding the sequence is obvious to the ordinary genomic scientist, the case should not affect claims to inventions identified by procedures that are not obvious at the time of patent application. Kubin does not call into question patents on DNA sequences that arise from genuine invention; rather it corrects the anomalously low threshold for nonobviousness established by Deuel. Kubin is not a “nail in the coffin of DNA sequence patents,” but rather a mechanism for culling marginal patents based on an accurate reading of the state of the science.