patenting DNA sequences

Phil Green pg at GENOME.WUSTL.EDU
Mon Oct 14 10:17:43 EST 1991

(1) Rob Harper writes:
>**********************  CLIP from New Scientist ********************
>New Scientist 7th Sept 1991 No 1785 page 22. 
>Title "Genome funds 'wasted' on patents.
>Walter Bodmer, the president of HUGO, told the BA that American
>researchers funded by federal agencies, including the NIH and DOE
>are filing patents for some 1000 partial sequences of human genetic 
>material each month. With each application costing $30, this amounts
>to some $30,000 a month. The initial cost of patenting is double 
>the cost of sequencing.

 Bodmer's suggestion that only American researchers are applying for
these patents is inaccurate and perhaps self-serving. The problem is
international. For example, it is widely believed that the major
British group engaged in partial cDNA sequencing are withholding their
data pending submission of their own patent application. The same
appears to be true of the French and Japanese cDNA sequencing efforts.
In this light, the outrage from these countries over Watson's
statement a few years ago, that information from the genome project
shouldn't be shared with anyone that didn't help pay for it, now seems
pretty hypocritical!

In most of these cases, I suspect the pressure to patent is coming not
from the scientists themselves but from their government sponsors, who
want to exploit any conceivable commercial advantage. But it is up to
the scientists to resist this pressure.  Otherwise, we could be
beginning an economic "World War 3" over the genome, of which the only
beneficiaries will be the lawyers (who are the functional and moral
equivalents of weapons dealers).

(2) Concerning the issue of patentability: The various national patent
offices have in the past already granted patents for the cloning of
several commercially important genes, so the question whether a
sequence is patentable seems to have been answered in the affirmative
(since having a clone is essentially equivalent to having its
sequence). In those cases, though, the function of the gene product
was known. The argument for patentability of partial cDNA sequences
seems to be that once one has an initial piece of a gene sequence, all
the rest, including figuring out what the gene product does, is
straightforward.  The partial sequence is supposed to be the
breakthrough which makes everything else possible. To most molecular
biologists that argument is ridiculous: determining the function of a
gene nearly always requires far more ingenuity and hard work than
sequencing it. (The only exceptions would be those cases in which one
can confidently guess the function from the sequence because of a very
strong homology with a gene sequences already in the database -- but
then the database sequences already provided the means to clone the
gene, so they, not the new partial sequence, represent the
breakthrough.)  If one is willing to dismiss the difficulty of finding
out the gene function, then why shouldn't making the cDNA library be
the patentable step, since the sequencing is straightforward?  In
fact, why not patent the genome itself, since making the cDNA library
is certainly straightforward?  Unfortunately, Patent Office expertise
in molecular biology seems to be low, and since in addition there is
likely to be considerable pressure coming from the governments
themselves I think we have to be concerned that the partial cDNA
patents may actually be granted.

(3) Bill Pearson writes:
>	I'm not sure where this $30 per patent figure comes from.  It is
>my understanding that it costs more like $50,000 to file a patent (lawyers
>fees, searches, who knows what).  I doubt whether many researchers are
>spending $50,000,000 a month on patent filings.

At least one of the prominent labs generating partial cDNA sequences
(at the NIH) is applying for patents not on individual sequences but
on groups of 1000 or so at a time. So the cost per sequence is
presumably roughly $50,000/1000 = $50. Bodmer's figure of $30 might
have been based on this sort of calculation, although the wording in
the New Scientist is garbled. Clearly, one way to put a stop to these
applications would be for the Patent Office to require a separate
application for each partial sequence.

(4) On the issue of "discovery" vs. "invention" -- since these partial
cDNA sequences are only 95-99% accurate, maybe they really are
"inventions" after all! 8-) 

Phil Green
standard disclaimer applies

More information about the Bioforum mailing list