Biotech patent questions/follow-up

Melody Ferguson (melody@molecumetics.com)
Tue, 28 Oct 1997 14:36:51 -0800

On Oct. 10, I requested information on searching for patents claiming
specific protein assays. The only reply came from Adrienne Shanler at
Cyanamid. She recommended the following for a prior art search
(Adrienne, I hope I have correctly represented you here. Please
correct me if I am wrong.):

1) First, Adrienne expressed sympathy :), then...

2) Check for synonyms for the protein including REGISTRY and enzyme
commission (EC) numbers. If cloned, check GENBANK and DGENE. She
noted that there might be several ECs, or they might be improperly
used.

3) Recommended terms such as: assay, method, quantify, quantitate,
method to detect, technique to quantitate (but watch out for false
hits from "method"). {I would add screen or clone, and maybe measure
(Xn) inhib?.}

4) If possible, try limiting by organism, type of assay, etc.

5) For patents and literature, Adrienne checks for name, synonyms, EC
numbers, abbreviations. For patents, she adds functionality (in case
the patent does not name the protein specifically). For literature,
she checks LOTS of files (which I won't try to list).

Many thanks to Adrienne. I was quite impressed.



In my case, I was able to limit the search to patents only (thank
goodness!!!!!).

Here is my experience for the first round:

1) The protein is not always specifically named in the claim (e.g.
they might reference "the protein with the sequence in figure X").

2) STN was invaluable in this search. The REGISTRY file has lots of
synonyms and definitely increased my retrieval. I can tell you that
the name of the protein I was given was not mentioned in one of my
answers! (Thanks CAS.)

3) The additional indexing in CAPLUS and BIOTECHABS is also quite
helpful.

4) In USPATFULL, try searching for the protein in the title,
abstract, and claims and the "assay" terms in the claims. Combine the
sets.

3) Displaying the USPATFUL claims is really valuable. I used the
smartselect command to select all the US answers I could display.

4) From my conversations with a biologist, I have been reminded that
protein characterization is a difficult process. Scientists may
report they have a certain protein, but later learn they have another;
protein names change, etc. So the "name game" is important here.
(The double jeapordy of getting false hits and missing hits applies
here.)

I typically don't limit with classification codes. However, I am
considering it for this search. Does any one want to comment on the
pros and cons for this particular search?



I apologize for the delay. Thanks for the assist. And, of course, I
would welcome any further comments. (I can always learn more!)

p.s. And for the real mystery - where did all the time between Oct.
10 (when I asked the question) and Oct. 28 (today) go???


Happy Halloween, everyone.

----------------------------------------------------------------------
Melody M. Ferguson
Information Specialist
MOLECUMETICS LTD.
2023 120th Ave. N.E. Suite 400
Bellevue, WA 98005-2199

Phone: (425) 990-1323
FAX: (425) 646-8890
e-mail: melody@molecumetics.com
----------------------------------------------------------------------