From: Brad Dolan <bdolan@use.usit.net>
Date: Wed, 16 Aug 95 18:33:35 PDT
To: cypherpunks@toad.com
Subject: genetic software patents (fwd)
Message-ID: <Pine.SOL.3.91.950816212608.27837F-100000@use.usit.net>
MIME-Version: 1.0
Content-Type: text/plain




---------- Forwarded message ----------
 
     Sometime ago, I reported on a recent discovery of how gene sequences
can be used for mathematical optimization calculations.  The method,
developed by Leonard Adelman (the "A" of RSA) provided a way of solving
the Traveling Salesman Problem using gene sequences (interestingly, the
paper by Hopfield that kicked off the neural network revival also solved
the TSP).  Computable genes blurs even more hardware and software (along
with hardware/software codesign tools), but has yet to sink into PTO
statutory and prior art thinking.

    In any event, the August 1995 issue of Dr. Dobb's Journal, page 127,
has a nicely illustrated article titled "Biochemical techniques take on
combinatorial problems".  If you or your clients are interested in this
new form of computing, get a copy of the article.  Adelman's original paper
titled "Molecular Computation of Solutions to Combinatorial Problems" and
was in the November 11, 1994 issue of Science.

    While you are at it, get a copy of an article that appeared in the
July 27, 1995 issue of Nature, page 307, titled "Protein molecules as
computational elements in living cells".  It starts out "Many proteins in
living cells appear to have as their primary function the transfer and
processing of information, rather than chemical transformation of metabolic
intermediates or the building of cellular structures".  Yet one more blur
or "hardware" and "software".  Geneware?

    A final article to get is "Circuit simulation of genetic networks",
appearing in the August 4, 1995 issue of Science, page 650.  It proposes
to simulate genetic networks (such as the bacteriophage lambda lysislysogeny
decision circuit) using the tools found in hardware/software codesign tools.
With VHDL as a digital circuit language, and AHDL as a analog circuit language,
are we going to need GHDL as a genetic circuit language?

    It is not hard to imagine in the future some VHDL code being mapped into
software with standard translators, into hardware using cell libraries, and
into gene sequences using sequence libraries interface to EDA design tools.
Properly structured sets of claims under the Doctrine of Equivalents could
claim coverage in all types of domains (in fact, the claims could be generated
by the design tool itself using a "cell" library of claim clauses).
Alternatively entire libraries of medical journals suddenly become potential
software/hardware prior art.



     While practical use of the gene technique is years off, it does raise
the issue of how the PTO is going to handle the prior art aspects, when
gene technology now becomes software and hardware prior art (as soon as, if
ever, the PTO first solves the software-software prior art problem, and then
the software/hardware codesign prior art problem).  Scientists are blurring
the lines of computing much faster than the PTO can respond; for example, the
proposed software patent guidelines reflect none of the above developments.

     Together these three papers represent the cutting cutting edge of issues
that will be affecting the patenting world in the years to come. Read them
now to get a jump.