Text 660, 229 rader
Skriven 2004-11-05 05:58:00 av Tinyurl.Com/uh3t (1:278/230)
Ärende: Re: No Grace Period for M
=================================


> From: an588@freenet.carleton.ca (Catherine Woodgold)
> Catalysts generally increase the rate of some reaction.
> That suggests that the reaction might occasionally
> proceed even in the absence of the catalyst.

That's partly true. Part of the action of a catalyst is to speed up a
chemical reaction that would occur anyway. But another part of the
action is to specifically direct the reaction in a particular
direction. Especially in reactions that transform organic chemicals
(anything with several carbon atoms linked together), there are many
different ways a high-energy molecule can decay to release its energy,
either by itself or when bumping into various other chemicals. But a
catalyst would provide a "bridge" over one specific energy barrier
whereby crossing that particular barrier (in that particular direction)
is rendered easy while all other directions across all other energy
barriers are still difficult. Without the catalyst, that particular
reaction path might be so rare as to never occur in a million years
while the molecule usually decays via other paths, but with the
catalyst that particular reactoin path might happen within seconds and
dominate over all other possible reaction paths. Even if the reaction
rate across the catalyzed path is slow, taking hours, if all other
reaction paths are even slower, taking days, in the end (if you wait
those hours to see what happened) that one catalyzed reaction path will
dominate.

There may still be some very weak catalysts which speed up their
particular reaction direction just a tiny bit, not enough to exceed the
reaction rate via other paths. For most purposes here, we can ignore
them as just noise in the situation.

Even a simple metal ion can achieve dominance of reaction directions,
as we use them in industrial processes, so I expect with a wide variety
of organic molecules randomly created in the pre-life times, enough of
them would achieve dominating catalytic action that we can just ignore
those which don't.

> I think catalytic cycles like that occurred, but that they didn't
> last and weren't Life.

I agree. We can ignore cycles whereby the catalytic activity at any
step around the loop is so very weak. Let's concentrate our study of
catalysts which dominate the direction of chemical reaction for any
given input chemicals (reactants), and thus chains of catalyst begats
catalyst where each catalytic step in the chain dominates, so when we
talk of a catalytic chain closing back on itself we have a good chance
of fecundity greater than one around the cycle.

> Things that replicate, and even grow exponentially, but are not Life
> include:
>  -- fire

No, fire doesn't grow exponentially, because it requires a large bulk
of hot material just to pass the ignition threshold, and the surface of
a large bulk brows only quadradic (and volume grows cubic) over time.

But that would still be good enough if not for the threshold effect at
a more fundamental level: A single molecule, or even a few molecules,
of hot material, can't possibly heat nearby material past the ignition
temperature to where the fire "replicates". If there in fact is such a
large bulk of hot material as to pass the ignition threshold, after
fire burns a while and depletes its available resources (free oxygen or
some reduced material such as cellulose), the temperature drops below
the ignition temperature, and that instance of fire goes extinct.

This is totally unlike my proposed catalytic-cycle replicator, whereby
a single molecule has a chance of starting exponential growth, and ten
copies of a single molecule has very high chance of growing rather than
dying out, and hundreds of copies of the single molecule are virtually
certain to grow exponentially rather than die out. Furthermore, as
resources are used up, replication slows down, and spontaneous
destruction begins to catch up, but at no point does the number of
copies of the molecule drop below a threshold for activity. Even if the
number drops from macroscopic quantities back down to a few hundred
copies, even then exponential growth can start again whenever resources
become available.

The logic why fire can't survive in small numbers of molecular
instances is similar to the reason why altruism can't come into being
unless by some sort of magic a majority of the population already is
altruistic.

Also you're stretching the definition of replication (converting other
kinds of materials into more of like-self materials faster than
like-self materials can naturally decay) to the limit. The only
property of fire that's different from non-fire is that it's very hot,
which is a quantitive largescale measurement, not quite a property of
individual particles of the material. If you have isolated particles of
very hot material, they very rapidly, almost instantly, radiate their
heat to become not-so-hot material. Their conversion of not-hot
material into very-hot material is very very temporary except when
there are macroscopic quantities of the material concentrated in a
small region all (or most) simultaneously very hot.

>  -- crystal growth

That doesn't grow exponentially either. Also the catalytic effect of a
seed crystal doesn't direct the action to a specific direction of
change among many as a catalyst does. (Except in rare cases such as
alum where there are two kinds of crystals each of which catalyzes only
its own kind. Still it can't convert anything except that one
particular chemical into crystals, so it can't fill the oceans.) Also
it's a dead-end process, converting one particular chemical into
crystals, but not producing anything that can be a catalyst for
converting something else. (Even with the case of sodium iodide
crystals used to seed crystalization of water vapor, the environment
where sodium iodide crystals grow isn't the same environment where they
can seed water vapor crystalizing, so you don't get a chain of one
catalyst begatting another.) But I guess you'd argue that's moot since
the crystals already begat themselves and don't need to begat anything
else.

If we require actual chemical changes, rather than just a change in
state such as crystalization, then crystal growth doesn't qualify as a
replicator. That may be too restrictive. We need to think of something
that allows chemical reactions, or electronic data changes, etc. etc.,
where true reactions are occurring even if not chemical, yet forbid
simple changes of state such as crystalization.

>  -- steam in bubbles in superheated boiling water

That is just a change of state again, with the same problems as crystal
growth.

>  -- stuff precipitating from supersaturated solution

That, and also crystalization, depends on a gross quantitative
property, like fire, with a threshold effect. You need the macroscopic
properties adjusted just right so it's ready to be triggered by the
seed crystal or whatever but doesn't go over the edge all by itself
without the trigger. You can contrive such situations, but they're
unlikely to hardly ever appear in the prebiotic earth's oceans. Also as
soon as the solution is diluted, the balance shifts in the other
direction, and the crystals rapidly dissolve and the precipate rapidly
re-dissolves. There's no way single particles of such crystal or
precipate could diffuse through neutral water to reach a new place they
could infect.

> I think catalytic cycles tend to be in the same category.

Nope. A single molecule of any of the catalysts from a catalytic cycle
could drift away from the now-resource-depleted region where it had
formed, into a new region with plenty of resources, and begin
exponential growth again. A single molecule would have a chance of
that, and thousands of molecules all diffusing at the same time would
surely spark a new region of exponential growth. This is not at all
like fire crystals bubbles or precipates.

> To be Life, it would have to eventually mutate.

I covered that a few months ago when we were discussing lipid bubbles
with ecosystems of replicators residing on them. Basically a well
established catalytic loop might have not quite a single loop of exact
chemical species but some branching around the loop so that at each
point around the loop all the catalysts are of the same type but
differing slightly. If some of the catalysts include metal ions such as
iron or copper, or include side chains that don't affect the catalytic
activity very much, the specific metal ion or side chain could vary in
this way among each type of catalyst. So instead of C1 begat C2 begat
C3 begat C1, we might have (C1a + C1b) begat (C2a + C2b + C2c) begat
(C3a + C3b) begat (C1a + C1b). In an environment where a lot of metal
ion or side-chain material is present, and where there's some bias as
to which version of a particular catalyst is begat by which version of
the previous catalyst, the loop might split into C1a begat C2b begat
C3b begat C1b begat (C2b + C2c) begat C3a begat C1a, thereby doubling
the length of the loop.

If the catalysts are such that they tend to polymerize, it's even
easier to achieve spontaneous variation (i.e. mutation): Suppose the
original loop is C1a begat C2a begat C3a begat C1a. Now suppose at some
point C1a attaches to another copy of itself to form a dimer, and each
unit of that dimer catalyzes production of C2a, and the reaction to
make the second molecule of C2a proceeds faster than the first due to
influence of the first already present. So then instead of one unit of
C1aC1a begatting one unit of C2a which drifts away, C1aC1a begats one
unit of C2a which stays attached until the second unit of C2a is
formed, at which the C2aC2a breaks loose as an already-formed-at-birth
dimer. Then likewise C2aC2a begats not individual C3a but dimers
C3aC3a. Likewise C3aC3a begats C1aC1a dimers. So suddenly the catalytic
cycle has mutated from a monomer to a dimer cycle. This might proceed
to some maximal optimum size where efficiency is maximized but the
chains aren't long enough for shear forces to break them apart faster
than they form.

Now suppose one of these C1a or C2a or C3a occasionally catalyzes the
production of a different flavor of the next in the cycle, so at some
point instead of C2aC2aC2aC2a we have C2aC2aC2bC2a. Now if the same
statistics as before comes into play, where the particular variation of
a catalytic class begats more of one variation than another of the next
class in the cycle, we can easily have alternating generations of the
cycle, which in effect doubles the cycle.

In short, with the catalysts in the cycle being a little non-specific
as to which version of the next type of catalyst they produce (begat),
and with polymerization of the individual catalytic units, there's an
immense range of possible mutations that can occur. If in different
environements various flavors of these are better at replicating and
surviving than others, we can already have Darwinian evolution going on
as these catalytic-loops adapt to various local environments.

> It seems more plausible to me that a strand of RNA
> catalysed the generation of a complementary strand.

How often does a strand of RNA spontaneously form out of Miller-Urey
gunk? Would a Miller-Urey gunk situation contain just the right mix of
chemicals that would support RNA spontaneously catalyzing its
complementary strand? My guess is that both answers are in the
negative, hardly ever such a RNA strand forming, and almost surely not
the right chemicals to support RNA replication. But enlighten me if you
know otherwise.

I think it's more likely that some random replicator came first, then
it polymerized, which yielded a very primitive genome, and later it
evolved to produce enzymes, and later one of those enzymes aided
replication by pattern, and later the original polymer was replaced by
RNA or DNA or polypeptides, and if not DNA then later it was replaced
by DNA. At each replacement event, the parasitic replacement out-bred
its parent species, driving it to extinction, removing all evidence of
the parent that we could see today, so the only evidence we see today
(including the fossil record for the past 3500 million years) is the
final generation in this sequence, the DNA generation.
---
ū RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info@bbsworld.com

---
 * RIMEGate(tm)V10.2á˙* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 11/5/04 5:58:09 AM
 * Origin: MoonDog BBS, Brooklyn,NY, 718 692-2498, 1:278/230 (1:278/230)