Gene Sequence

I'm thinking that this is the type of thing I see now and again as a trivial name---an example would be patent 6,107,080 (a biosensor) where a sequence listing is part of the specification, but because it isn't a separate section of the patent and not in the claims, the sequence probably would not be indexed. The claims specify several generic gene products like streptavidin complex, and are characterized by their function (binding). I pulled this one at random, but see sequences often enough that have to do with diagnostic devices. The spec says in part:
...the ligand in the biosensor is anchored to biosensor surface, i.e., embedded within the hydrocarbon-chain monolayer, by a coiled-coil heterodimer complex formed of two subunit peptides. The heterodimer-subunit peptides employed in the biosensor invention are two non-identical, preferably oppositely charged polypeptide chains, typically each about 21 to about 70 residues in length, having an amino acid sequence compatible with their formation into two-stranded .alpha.-helical heterodimeric coiled-coils. They are designated herein as HSP1 (heterodimer-subunit peptide 1), and HSP2 (heterodimer-subunit peptide 2). In the discussion below, HSP1 will refer to the peptide attached to the biosensor surface in the biosensor, and HSP2, to the peptide having an attached ligand. It will be understood that these designations refer to the functional role played by the subunit peptide, not the actual peptide sequence.

I pulled this patent at random, but see partial sequences often enough that have to do with diagnostic devices. The device world aims at getting the most generic identification possible---so if a competitor wants a rapamycin effect, my patent claiming antibiotics trumps. In the above patent, the sequence is given in order to enable, not to claim it. You often see "kitchen sink" listings of common chemical names in order to teach, so that your competitors can't design around you with an obvious substitute.

There are changes on the horizon on indexing protocols, but for now I'm assuming if a separate document/ attachment exists for a sequence listing, or of course if it's been deposited in a public database, that's how it's done; they aren't searching for little pieces of sequences within patent specifications. I went to one semantic search seminar in Napa where the guy proposed using noun-verb-noun phrases to parse claims---he looked a little stunned when I told him a claim didn't need a verb.

Is this the kind of thing you were driving at? For a test, take your compound common names and look for them under the umbrella of usage platform---diagnostics, treatment are huge areas that can de mined.

Let's hope Rob Austin or other expert will enlighten us further on whether "collateral" sequences get collected in indexing.

I think it is possible to claim a gene sequence by its name / ID rather than indexing its genetic sequence (in claims). I found some recent US patents/applications of this type , a few examples (by random search)  are given below:
 
20070157329 , 20050260581, 20070178114 , 7419807 , 7465851
 
According to my understanding, sequence search in Genome quest / CAS Registry Blast would retrieve references that has not claimed but indexed the sequence in any part of the disclosure.

Are you interested in knowing  a patent claiming for a gene sequence by its name / ID in which the sequence is not indexed in any part of the document?

We have come across patent claiming inventions related to HPV 16 E6 sequence by referring to the name only.