Prev: Prime Number Fluid model data and graphs
Next: Just One Good Cipher Will Do.
From: Mok-Kong Shen on 22 May 2010 12:17
Dave -Turner wrote:
> To expand on that - http://en.wikipedia.org/wiki/Leonard_Adleman
>
> In 1994, his paper Molecular Computation of Solutions To Combinatorial
> Problems described the experimental use of DNA as a computational system. In
> it, he solved a seven-node instance of the Hamiltonian Graph problem, an
> NP-complete problem similar to the travelling salesman problem. While the
> solution to a seven-node instance is trivial, this paper is the first known
> instance of the successful use of DNA to compute an algorithm. DNA computing
> has been shown to have potential as a means to solve several other
> large-scale combinatorial search problems.
>
> In 2002, he and his research group managed to solve a 'nontrivial' problem
> using DNA computation. Specifically, they solved a 20-variable 3-SAT problem
> having more than 1 million potential solutions. They did it in a manner
> similar to the one Adleman used in his seminal 1994 paper. First, a mixture
> of DNA strands logically representative of the problem's solution space was
> synthesized. This mixture was then operated upon algorithmically using
> biochemical techniques to winnow out the 'incorrect' strands, leaving behind
> only those strands that 'satisfied' the problem. Analysis of the nucleotide
> sequence of these remaining strands revealed 'correct' solutions to the
> original problem.

I am an absolute layman in that field. What I guess is that up to now
the scientists had severe constraints in what they could obtain, i.e.
they might only be able to tweak a little bit from what is present in
nature, while now (I surmise) they are relatively free to plan/design
and generate what they want.

M. K. Shen
From: Maaartin on 22 May 2010 13:47
On May 22, 5:46 pm, "Dave -Turner" <ad...(a)127.0.0.1> wrote:
> In 2002, he and his research group managed to solve a 'nontrivial' problem
> using DNA computation. Specifically, they solved a 20-variable 3-SAT problem
> having more than 1 million potential solutions.

It's nice that it works, but I'm afraid, they're come 20+ years too
late. Sure, there will be some improvements there, but I don't thing
they can ever catch up the standard technology. Compared to the
switching speed of a CPU, the DNA moves quite slow, there's a lot of
parallelism, but you can use many CPUs for all problems solvable by
DNA computing.

A 20-variable 3-SAT may sound non-trivial - for a paper and pencil
solver. Moreover, finding a solution for an NP problem is one thing
(you can guess and with many guesses in parallel you simply must find
it), proving that no solution exists is quite different, and I can't
imagine, how this could be using a DNA computer.

While I consider the synthetic bacterial cell to be a great
achievement with huge potential in biology, medicine, and industry, I
don't think it'll *ever* provide any useful computing power. However,
I really don't know much on the subject, and I'm curios if somebody
can prove me wrong.
From: Casper H.S. Dik on 22 May 2010 14:33
Mok-Kong Shen <mok-kong.shen(a)t-online.de> writes:

>I am an absolute layman in that field. What I guess is that up to now
>the scientists had severe constraints in what they could obtain, i.e.
>they might only be able to tweak a little bit from what is present in
>nature, while now (I surmise) they are relatively free to plan/design
>and generate what they want.

It's to analyze and reproduce it; but we still don't know exactly
fully how the cell and the DNA inside it works. Until we know that,
we cannot make living new creatures from scratch. (But, unfortunately,
it does make it possible to create smallpox if you have a DNA sequence.

smallpox is, of course, a very complex virus, but I believe they've done
this with polio.

Casper
--
Expressed in this posting are my opinions. They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.
From: Mok-Kong Shen on 22 May 2010 14:58
Casper H.S. Dik wrote:

> It's to analyze and reproduce it; but we still don't know exactly
> fully how the cell and the DNA inside it works. Until we know that,
> we cannot make living new creatures from scratch. (But, unfortunately,
> it does make it possible to create smallpox if you have a DNA sequence.
>
> smallpox is, of course, a very complex virus, but I believe they've done
> this with polio.

The webpage I cited is exteremely terse on that and says nothing on
exactly what has been achieved. On the other hand, if the cell that has
been sythetized couldn't live and reproduce, one could hardly regard
the research work as any (significant) success, I would think.

M. K. Shen


From: adacrypt on 23 May 2010 02:06
On May 22, 7:58 pm, Mok-Kong Shen <mok-kong.s...(a)t-online.de> wrote:
> Casper H.S. Dik wrote:
> > It's to analyze and reproduce it; but we still don't know exactly
> > fully how the cell and the DNA inside it works. Until we know that,
> > we cannot make living new creatures from scratch.  (But, unfortunately,
> > it does make it possible to create smallpox if you have a DNA sequence.
>
> > smallpox is, of course, a very complex virus, but I believe they've done
> > this with polio.
>
> The webpage I cited is exteremely terse on that and says nothing on
> exactly what has been achieved. On the other hand, if the cell that has
> been sythetized couldn't live and reproduce, one could hardly regard
> the research work as any (significant) success, I would think.
>
> M. K. Shen

Given the myriad of potential models that already exist in the whole
of physics and mathematics surely it is anamolous to jump on this one
in medical science as having extraordinary potential to cryptogology
and for what exceptional reasons? - the statistical experiment of the
DNA probability?

There is nothing to pursue in this from a cryptography point of view
in mybelief.

This is not a crypto goalpost in any way.

I believe in looking under every stone for new research areas in
crypto research but this is not one - I condone the attitude of trying
anything at the same time - The industry has been barking up the wrong
tree for some time now - ever since crypto became intensely number-
theoretic with the arrival of computers in fact - it crying out for
change.

The stupidity of complexity-theoretic research is nothing more than a
distant, obsolete piece of history that the no-hopers will not allow
to go away - it is their only cant in posts to this news group.

It strikes me that you are mixing the jerseys here a bit.

The scientic achievement by the scientists at the J. Craig Venter
Instuitute is marvellous but projecting it as a potential field of
research in cryptography on the back of the tenuous fact of the
probability aspect of DNA is ridiculous.

I still say the principle of trying is right, however outrageous it
may seem - that is research but it would playing to wrong goalposts
here entirely . - adacrypt
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: Prime Number Fluid model data and graphs
Next: Just One Good Cipher Will Do.