[118709] in Cypherpunks
DNA computing - an explanation before you get completely carried away
daemon@ATHENA.MIT.EDU (Jim Choate)
Wed Oct 6 03:43:24 1999
From: Jim Choate <ravage@einstein.ssz.com>
Message-Id: <199910060731.CAA27496@einstein.ssz.com>
To: cypherpunks@einstein.ssz.com
Date: Wed, 6 Oct 1999 02:31:21 -0500 (CDT)
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Reply-To: Jim Choate <ravage@einstein.ssz.com>
Ok, enough name calling over this worthless piece of mis-direction on
anonymous' part. Unless that's the reason you're actualy here, in which case
have at it.
Here's the whole story....
Somebody on coderpunks asked how to create a time-released envelope for
storing information in.
I made the comment that DNA computing techniques could be used to build a
TIMER (note: NOT an actual encryption system in DNA) that was reliable and
could use some recent techniques for tagging small quantities of resultant
on glass plates.
For some unknown (and irrational) reason several people on the list got
stuck in the "But to make a DNA computer takes so much material". This
argument is irrelevant and immaterial. We are NOT building a crypto system
cracker in DNA. We don't need large quantities of DNA to compute individual
potential keys.
What we are doing is building a timer that releases two proteins at some
designated time. The first protein is only a tagant that says "Yes, we have
reached our desired output stage". The second tagant would be the actual *key*
to the external crypto system, again, not the crypto system itself. Only a
numeric key that is encoded in the sequence of the DNA which can be decoded by
examining the protein string resultant. Conceptualy I see breaking the
actual key into short pieces, say 3-pair codons, and then hiding them in
'trash' sections of the DNA sequence. As the DNA program is run it would
collect and collate a sequence of vectors to the key sections. Only at the
appropriate time would the vectors come together and enable the DNA to
release the key.
I also commented that there was an interesting article on the SciAm webpage
about RNA and rhibosomes that was relevant.
I also made the observation that by hijacking the apoptosis mechanism in a
cell like yeast an even better and more efficient timer could be constructed
using something like circadium rythms.
The point being that this sort of mechanism could be built in a coffee cup
if one desired. Whereas anonymous and his cohorts kept blabbering about the
amount of DNA required and such. The entire line is irrelevant to this
issue. Apparently their mental wonderlust was insufficiently fed on
coderpunks so they exported the whole thing over here.
Put 10^8 or so DNA strands in there (to build a sufficient concentration to
make it reasonably easy to detect) and sit back and wait for the taggant to
pop out. You could probably do it in a 100ml Erlynmyer.
At least one person brought up the dependency of chemical reactions to
temperature. DNA is perfect for this, if it's too cold it runs slow and if
it's much above body temperature the reaction doesn't run faster but rather
the DNA comes apart and no reaction occurs at all.
The only hole in this is if somebody could take the DNA strand (my example
used 10,000 base pairs) and work through its programming via a computer or
other method faster than the actual chemical reaction. Speed is the one
major drawback to DNA computations.
A personal observation:
There are times where I wonder if certain parties on this list (as well as
other controversial ones) don't intentionaly de-rail discussions by practicing
these sorts of tactics. I hope it's just their own immaturity and not as some
schill (well really I hope it's rampant paranoid imagination on my part).
____________________________________________________________________
The best lack all conviction, while the worst are full
of passionate intensity.
W.B. Yeats
The Armadillo Group ,::////;::-. James Choate
Austin, Tx /:'///// ``::>/|/ ravage@ssz.com
www.ssz.com .', |||| `/( e\ 512-451-7087
-====~~mm-'`-```-mm --'-
--------------------------------------------------------------------
