Re: DNA carries information

On 27 Aug 2006 16:41:37 -0700, "Friar Broccoli" <EliasRK@xxxxxxxxx>
wrote:

r norman wrote:
On 27 Aug 2006 12:00:22 -0700, "Friar Broccoli" <EliasRK@xxxxxxxxx>
wrote:

r norman wrote:
On Tue, 22 Aug 2006 23:54:44 +1000, j.wilkins1@xxxxxxxxx (John
Wilkins) wrote:

r norman <r_s_norman@xxxxxxxxxxxx> wrote:

On Tue, 22 Aug 2006 01:38:33 GMT, "Perplexed in Peoria"
<jimmenegay@xxxxxxxxxxxxx> wrote:


"Friar Broccoli" <EliasRK@xxxxxxxxx> wrote...
r norman wrote:
On Sun, 20 Aug 2006 17:23:06 GMT, "Perplexed in Peoria"
<jimmenegay@xxxxxxxxxxxxx> wrote:

Furthermore, there is an intimate formal
connection between Shannon information and negative entropy, at least
entropy in the statistical thermo sense, not the classical heat-based
thermo.

Not trying to pull this phrase out of context. Just pointing to what
I took to be the core point in an article I very much appreciated.

Well, since this sentence has been highlighted, let me point out
that there is also a connection with classical heat-based thermo.
It is sketched in Pierce's classic popularization of Shannon
"Symbols, Signals, and Noise". It involves a novel form of heat
engine utilizing a gas consisting of a single molecule and a
demon who can recognize which side of a frictionless piston
contains the molecule, but must communicate this information
(to a second demon controlling a valve) over a Gaussian channel
at temperature T. In this case, the demons can't beat the
second law of thermodynamics.

Maxwell's Demon is a long recognized device. Usually it consists of
the demon controlling a door between two gas filled chambers. When a
high velocity molecule approaches the door from one side, the demon
opens it, but closes it for low velocity molecules. It does the
reverse on the other side. So all the high velocity molecules end up
on one side, the low velocity ones on the other, producing an
imbalance of temperature (and pressure) which can be used to drive an
engine.

The standard demon is often exorcised by the fact that opening and
closing the door itself requires energy. If not and the door is in
thermal equilibrium, it will simply flutter open and closed on its
own. The informational exorcism relies on the fact that the demon
must acquire some "information" from the oncoming molecule by
measuring its position and velocity. That disturbs the molecule,
altering its properties. For example, if you shine a light on the
molecule to see it, the interaction between photon and molecule will
cause it to deviate in its path. My impression is that Brillouin
("Science and Information Theory") described the information theory
aspect and perhaps actually developed that concept. Brillouin's first
edition predates Pierce.

.

Brioullin showed that the employment of information requires
energy but he did not show that information and energy are
interconvertible (in fact memory tells me he said the
opposite).

You are right. I tried to be careful not so say that
information and energy are convertible, didn't I?


OK, I was going to try and have one of my handwaving fights
with yous guys ober this, but then reread this old POTM which
was (indirectly) nominated by the fat philosopher (who maybe
wasn't then). See one of:

http://www.talkorigins.org/origins/postmonth/oct00.html
http://groups.google.com/group/talk.origins/msg/8ee7672dcd41b6a2


There is a formal relationship between information and entropy but
that does not carry into a physical equivalence.

I think Davisson rather convincingly puts this idea to bed.
There is a physical equivalence.

His article is admittedly a bit heavy, but I can now make my
way through it with (what I can convince myself is
understanding), so if you guys really want a food fight on this
issue, I could write up a simplified description of (what I
think are) the key features of the demon/door IPU in this case,
together with a simplified argument for a direct physical
relationship between information and entropy.

And in another post I did try to say that the physical
manifestation of information in the real world of matter and
energy does require work/energy to store and to communicate
the information. Actually, my memory seems to tell me that it
is possible to store and retrieve information as a
"spontaneous" energy-down-hill (entropy increasing) process
but it takes energy to erase the storage so you can use it
again.

This is dealt with at length by Davisson.

The physics of information is a fascinating subject, but rather far
afield from whether or not it is useful to say that DNA carries
information.

Against sane people yes, but Wilkins is trying to assert that
information is not a physical property, like colour, but an
abstract one like number. Since DNA is encoded physically, I
need to undercut this little bit of philosophical mayhem as a
basic (though not conclusive) step in tying physical DNA to
physically encoded information.

[closing comment snipped for no particular reason]


I certainly have no interest in pursuing the Maxwell Demon story
because I don't think there is any problem with it. In order to
function, the demon must interact physically with the molecules it
observes and with the door. These interactions are not taken into
account in the naive version that is a perpetual-motion machine or an
entropy decreaser. Whether you describe the interactions as
information or as energy or as "you cannot measure something with
disturbing it" is irrelevant; they will end up being equivalent.

This "cannot measure something without disturbing it" business
was dealt with (at least indirectly) by Davisson. One
possibility is that fast moving particles radiate
electromagnetic radiation at a slightly different frequency
than slower moving particles, thus permitting information
collection (velocity, location and direction) with close to
zero energy loss to the system.

But this is really beside the point. The purpose of the demon
in this particular context (ours here) is to ask the question:
What would happen if the demon itself could perform all the
other operations (detection, door opening and closing etc.)
without using any energy, so that we can focus exclusively on
the information use/storage problem. This thought experiment
shows that the information storage (or eraser) by itself
defeats the demon's effort to violate the second law.

Thus, yes, there are complicating factors, but the information
is real and has a real thermodynamic equivalence.


I still maintain that John is (and I am) correct in saying
that the formal description of information is a mathematical
construct.

You can make a mathematical construct of anything, but the fact
of that construct does not remove the underlying reality of the
thing or property (emergent or otherwise) modelled. So the
fact of a construct establishes nothing.

It is a measure on a set.

Just to confuse things a bit more. A measurement is a
procedure. Procedures must exist somewhere (in your head as a
neuron pattern, or in a computer etc.) So this information
clearly exists somewhere. If it didn't exist (wasn't encoded)
SOMEWHERE it couldn't be info., because among other things
there would be NOTHING to transmit.

You can measure the Shannon information in a sequence of
characters by thinking about the set and doing calculations in
your head or on a computer or on a piece of paper. Counting
bits of information in DNA by looking at the gene sequences is
just such a task. The records stored in all the
bioinformatics databases are just such records and the
"information" they contain is quite independent of the fact
that real sequences are made of DNA and require specific
enzymes and base pair bonding and all to function. In this
sense, Shannon information can be related to statistical
thermodynamic entropy, which can also be defined as an
abstract mathematical measure of a particular subspace of a
phase space.

As I see things you have confounded information in DNA as held in
and "read" by a cell with information about DNA held by
computers (which is substantially different). Then you have
performed an operation on the computer data, apparently
implying that that operation (which is a coded procedure) is
not real. In short, your final assumption (that the final
measurement is abstract) is false and it would be irrelevant
even if it were true.


It is when you actually build an Information Processing System
(IPS) to function in the real world of matter and energy,
space and time, that you end up equating information and
energy. And claiming that DNA carries information because it
is manipulated by an IPS is rather chasing your own tail.

Yes, as far as I can tell, I am doing the same thing John is
doing (using circular logic), by assuming that information is
information because it is manipulated by an IPS, without
providing a coherent independent definition of an IPS.

If I am not overcome by laziness ((a likely outcome) (and if
that doesn't get me, stupidity surely will)) I will try to
address this and other weaknesses by linking/defining
information as 'procedures that result in efficiencies'.

I believe that John's point is that, when discussing energy,
it is better to talk about the biophysics of the molecules and
their interaction and when talking about sequences it is
better to talk about specificity of binding. Counting bits or
saying that "information increases because of evolution" or
whatever you do say that is quantitative and makes use of a
particular metric of information is likely not to be useful. I
only add -- not at this time. Just you wait 'enry 'iggins!

I agree, but as previously stated, what we mean by information
is (it seems to me) well captured by applying efficiency
measures to the procedures (including cell procedures)
information encodes. So I agree, Shannon definitely is not
enough to capture this.

But of course, I can't do the math, so who do I thing I'm
fooling. :-(.


OK, so you made me actually read Davisson's POTM.

Right after his "first conclusion" he states "under realistic
conditions, the information contribution to thermodynamic entropy is
so small that it can be safely ignored". That is a point I made in my
post back in the closing comment you snipped: "Certainly the ATP spent
in replicating, transcribing, or translating a nucleotide sequence far
exceeds the theoretical physical energy requirement. And when you
count the ATP spent in producing the elaborate cellular machinery
necessary to do all that "information processing" (a phrase
deliberately chosen for lack of a better choice), the theory of
"energetics of information" certainly is irrelevant."

In other words, it is completely irrelevant whether or not information
is connected to thermodynamics and energy. All real information
processing systems gobble up and dissipate so much energy that the
amount associated with information is absolutely trivial. It would be
like going to a jewelry store and worrying about the carat weight of a
diamond because you know that its mass must be an integral multiple of
the mass of a carbon atom. It would be like worrying that your
electric bill might be inaccurate because it doesn't take into
consideration the fact that energy is quantized. The thermodynamic
energy of information is trivial. Of course there are exceptions:
quantum computing or trying to account for conservation of energy in
the universe when information falls into a black hole. That is why I
said "The physics of information is a fascinating subject, but rather
far afield from whether or not it is useful to say that DNA carries
information."

As to Maxwell's demon, all I was trying to point out was that Pierce
was not original in arguing for an informational exorcism. I
completely forgot about Szilard but I certainly would have remembered
had I been able to browse the Brillouin book I mentioned that is
somehow misplaced on my shelves (but, then, everything is misplaced).

You say "The purpose of the demon
in this particular context (ours here) is to ask the question:
What would happen if the demon itself could perform all the
other operations (detection, door opening and closing etc.)
without using any energy, so that we can focus exclusively on
the information use/storage problem. ". You have just thrown out the
baby with the bath water. The energetics of information is exactly
connected with the physical act of detecting the particle, determining
what state the door is in, opening and closing the door and making
sure that it stays in the desired position even though it is bombarded
by thermal agitation (being struck by all those molecules it is
herding). The fact is that as soon as one side of the chamber reaches
a slightly higher temperature/pressure than the other, there will be a
directed force on the door that must be resisted and opening and
closing it would take even more work. I argue that the energetics of
detecting the particles and controlling the door are equivalent to the
informational interpretation. You don't need to invoke information as
such to exorcise the demon. You don't need to invoke information as
such to do molecular biology.

The major point here is that "information theory" is not a necessary
part of the explanation of how the Maxwell Demon system works but it
is a very useful way for us to approach and analyse the system and
understand it. Again to try to paraphrase John, "information theory"
is not a necessary part of the explanation of how molecular biology
and evolution work and it is not even clear that it is a useful way
for us to approach and analyse the system.

I say that information is a "measure on a set". You respond " A
measurement is a procedure ...". I think you don't appreciate the
rather subtle point I was making that "measure on a set" is a
technical term in mathematics. See Wikipedia, "measure theory"
http://en.wikipedia.org/wiki/Measure_theory
which says "In mathematics, a measure is a function that assigns a
number, e.g., a "size", "volume", or "probability", to subsets of a
given set.". It is the connection to probability theory that ties
measures to information and entropy. In that sense, Shannon
information is a pure mathematical abstraction rather akin to "number"
and not a physical property like "color", to use the example your
cited.

You say "you have confounded information in DNA as held in
and "read" by a cell with information about DNA held by
computers (which is substantially different). " That is exactly what
I did because I was trying to show two completely different concepts
of information. One of them, the computer processing type, is a pure
mathematical concept with no connection to physical energy. The other
is the expression of information in a physical system which does
require energy and ties information to thermodynamics. There are
mathematical systems and there is physical reality. Very often we can
create mathematical systems which seem to closely correspond to
reality but they are very different concepts philosophically. The
mathematics is not the reality (unless you are a string theorist).

The problem still remains: it is intriguing to think in terms of
information theory metaphors when dealing with genetics and evolution.
Try to pin down exactly what you mean by defining your words and
making definitive statements that correspond to physical reality and
that can lead to experimental test and you immediately run into
insuperable problems. So where does that leave the information
metaphor?





.

Relevant Pages

  • Re: DNA carries information
    ... entropy in the statistical thermo sense, ... demon who can recognize which side of a frictionless piston ... contains the molecule, ... closing the door itself requires energy. ...
    (talk.origins)
  • Re: DNA carries information
    ... entropy in the statistical thermo sense, ... demon who can recognize which side of a frictionless piston ... contains the molecule, ... closing the door itself requires energy. ...
    (talk.origins)
  • Re: DNA carries information
    ... entropy in the statistical thermo sense, ... demon who can recognize which side of a frictionless piston ... contains the molecule, ... closing the door itself requires energy. ...
    (talk.origins)
  • Re: DNA carries information
    ... entropy in the statistical thermo sense, ... demon who can recognize which side of a frictionless piston ... contains the molecule, ... closing the door itself requires energy. ...
    (talk.origins)
  • Particle Self-Energy and the Transition Zone
    ... Particle Self-Energy and the Transition Zone: ... the new physics of the 21st century. ... The universal principle of energy discovered in the ... the fractional electric charge exhibited by quarks, ...
    (sci.philosophy.meta)