|
Author
|
Topic: Dermott J. Mullan: Probabilities of randomly assembling a primitive cell on Earth
|
charlie d.
Member
Member # 159
|
posted 28. December 2002 23:56
But that's entirely my point, Dermott! Your manuscript makes up a strawman model (random assembly of cells in one single swoop) that nobody - by your own admission - argues for in the scientific literature, then goes through a whole set of profound-looking calculations to conclude that for that to happen design would be needed. Well, duh.
I could make a similar argument that intelligent design is involved in diamond formation, because carbon atoms are just very unlikely to fall all in the right place, randomly, in one swoop. In fact, intelligent design would be needed for any dumb rock to self-assemble randomly. And if one pointed out to me that we in fact have some ideas about how diamonds or rocks formed, and that they don't assemble randomly, I could suggest that perhaps in light of my work one should consider random assembly and design as a valid alternative to existing models.
What this kind of argument adds to the scientific understanding of OoL, or even the intellectual exploration of ID, completely escapes me. Anyway, that's it for me here.
[ 28. December 2002, 23:57: Message edited by: charlie d. ]
IP: Logged
|
|
Dermott J. Mullan
Member
Member # 583
|
posted 06. January 2003 21:35
reply to charlie d (message of December 28)
Your remarks about an analogy between diamond formation and the assembly of the first cell are (I believe) not relevant. I believe that there are real distinctions between assembly of DNA (or RNA) and assembly of a crystal. Why?
Because there are well defined physical laws which control what happens in a phase transition. Near a critical temperature, it is well known that the correlation lengths increase greatly, and an ordering parameter of some sort takes on values that set the stage for a new phase to become dominant.
If the new stage is a crystalline solid, the laws of formation are known: the forces which exist between (say) the carbon atoms in a diamond crystal can be quantified, and then one can examine how the Gibbs free energy of the solid phase can be minimized under any given set of conditions. In a supercooled fluid, a perturbation with periodic geometry is unstable and therefore tends to form a regularly spaced lattice. The size of the lattice is determined by the interatomic potential. With carbon, the atoms find it more favorable to arrange themselves as graphite in some situations, and as diamond under other conditions. The fine details of interatomic forces determine what the crystal structure will be in any given situation.
Is there any role for the action of random assembly when a diamond crystallizes? Yes, to the extent that a random C atom is located in a particular site in the lattice. But the formation of the lattice structure itself is not at all random. This structure arises of necessity from the laws of physics.
However RNA (or DNA) is not assembled as the result of a phase transition. There are no intermolecular forces which ensure that (say) only the combination AAU in a nucleotide will have minimum Gibbs free energy. On the contrary, the structure of an RNA molecule, with its sequence of triplet AGA AAU GCU UAC etc etc cxan hardly be determined to any significant extent by the laws of intermolecular forces: otherwise, the same pattern of triplets would show up in all DNA molecules.
And the genetic code is not based explicitly on intermolecular forces either. There is no reason (from an intermolecular force point of view) why (say) the triplet CUC must lead to the attraction of leucine and only leucine when the mRNA goes out into the cell looking for an amino acid to insert into a new protein. In fact, in mitochondria, the same triplet CUC encodes for threonine.
I conclude that the physical processes which are at work in leading to a particular crystalline form of a solid are simply not relevant to the process of putting together the first RNA.
The problem still is: how will a strand of cellular RNA be assembled in such away as to be able to encode for all the proteins that the cell needs to exist? There are no physical laws which ensure that the triplets of codons will arrange themselves in just the right order. So why not see how this might be achieved by random assembly?
What I have done is to show that random assembly of atoms and molecules CAN lead to the first cellular RNA if we concentrate on the right regions of parameter space. If it can be shown that my conclusions are correct, i.e. that random assembly CAN do the task of assembling the cellular hardware, then I think that this a positive step in the OoL problem: to the extent that IDT may have something to contribute to the appearance of the first cell, my conclusion suggests that IDT should be concentrating its efforts not on assembly of the hardware, but elsewhere.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 07. January 2003 11:07
Dermott:
I see, you are one of those intransigent Gibbsists, who insist claiming a hypothetical "free energy", that nobody has ever seen, makes things fall into place, just like that. Of course, there is no evidence of intermediates in the process (has anyone ever seen atoms in a half-formed crystal?). Besides, free energy is clearly tautological: why is a conformation stable? Because it minimizes free energy. But why does it minimize free energy? Because of inherent stability. See the point?
Jokes aside, you misunderstand me. I am not claiming that there is an analogy between the design hypothesis as it applies to OoL and to crystal. Of course there isn't.
I am saying that if one makes a strawman of OoL hypothesis (random cellular assembly, which is a non-existent hypothesis, scientifically speaking) to argue for a design inference, like you do, one might as well make a strawman hypothesis of crystal formation (random atomic assembly) and argue for intelligently designed crystals.
If you started discussing OoL with the same scientific attitude and knowledge you apply to crystal formation and entropy, you'd realize "random cellular assembly" is not an option on the scientific table, so whether you can use it to argue design is irrelevant.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 07. January 2003 20:47
Charlie,
Dermott is in fact addressing modern thought concerning the OOL. Natural selection cannot select that which does not exist. Which is why many many mainstream theories of OOL do posit a completely random formation of the first cell.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 07. January 2003 22:17
Nelson
quote:
Dermott is in fact addressing modern thought concerning the OOL. Natural selection cannot select that which does not exist. Which is why many many mainstream theories of OOL do posit a completely random formation of the first cell.
Perhaps you could provide us with these many mainstream theories of OOL that posit a completely random formation of the first cell.
As far as I can tell most 'mainstream theories of OOL' seem to suggest some precursors to DNA such as RNA world, PNA world
or perhaps this link
quote:
Virtually all biologists now agree that bacterial cells cannot form from nonliving chemicals in one step. If life arises from nonliving chemicals, there must be intermediate forms, "precellular life." Of the various theories of precellular life, the most popular contender today is "the RNA world."
The Beginnings of Life on Earth
by Christian De Duve
The RNA World: The Dawn of Life
May we expect some references from you to support your claims (*)
(*) Changed per moderator request. I am nevertheless always interested in new materials and evidence.
[ 08. January 2003, 00:15: Message edited by: Frances ]
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 07. January 2003 23:37
Frances:
Comments like these are becoming more and more common:
"Looking forward to your references"
"Looking forward to your evidence"
I find these sorts of comments patronizing, and disingenuous. Please stop.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 08. January 2003 08:18
Nelson, not only it is entirely untrue that "many many mainstream theories of OOL do posit a completely random formation of the first cell" (in fact, AFAIK none does), but Dermott admitted as much in a previous post: quote:
Dermott:
How can I find a published paper in the recent biological literature about random assembly of the first cell if (according to your claim) the vast majority of modern bioogists do not believe in the process? The scientists I know do not write papers on topics they do not believe in.
Let's try to have a serious discussion, if possible.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 08. January 2003 20:15
Frances and Charlie,
I'm still digesting the paper, so I'll keep my comments breif until I really have a chance to sit down and read it. However, it is clear to me what his thesis is.
quote:
We evaluate the probability Pr that the RNA of the first cell was
assembled randomly in the time available (1.11 billion years
[b.y.]).
quote:
Our calculations refer only to the assembling of a cell in which
he genetic code is already at work. We do not address the origin
of the genetic code itself.
Now first of all, it is unclear to me what Dermott means by "random"; if he means anything that doesn't involve an intelligent designer, then his term "random assembly" could possibly include natural selection. However, if he is talking about randomly building a strand of RNA one base at a time using whichever base is nearby, then I think his paper is relevant to the first appearance, which the RNA world hypothesis does include in it's model.
Some quotations provide some clues on whether he is addressing natural selection in his calculations for example:
quote:
The change of even a single amino acid in a chain of dozens or hundreds of amino acids may in certain cases disrupt the functioning of the protein.
and
quote:
Therefore, if the first cell required (say) 30 proteins to become operational, the chance of assembling its RNA at random in the primeval soup after 1065 collisions is less than one in 1047m+147. The exponent in this result rapidly becomes large even if we allow for only marginal specificity. For example, if m has a value of 2,
Pr is less than one in 10240. And if m is set equal to its average value mav = 9, Pr falls to less than one in 10570.
This is extremely relevant to current OOL. Using only a random pool of RNAs , RNA world workers try to select for RNAs that perform a certain highly specified function. In the RNA World hypothesis, RNAs are generated by a chance process, followed by their selection based on a necessary interaction or function.
What I think accentuates his paper is that if the early earth consisted of a "soup" of a variety of different organic compounds, the early life forms would have to be able to synthesize their own components nearly from scratch. Not just any RNA strands would work. The RNA would have to be able to extract or build up the components from which it is made, as well as be able to catalyse its replication.
Since natural selection requires function to operate, I would think that some of these calculations are relevant. Since function comes about by pure chance, natural selection needs the random formation event to occur, and this is highly improbable.
[ 08. January 2003, 20:20: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 08. January 2003 23:57
Nelson
quote:
Since natural selection requires function to operate, I would think that some of these calculations are relevant. Since function comes about by pure chance, natural selection needs the random formation event to occur, and this is highly improbable.
You seem to be confusing randomness and unpredictability. Let's see if I can clarify this: mutations and variation in the genome is random with respect to its environment, selection will retain those variations which provide for a positive advantage and eliminate those with a negative advantage.
Random events are highly probable, a specific random event is however far less probable. There is no NEED that a particular event occurs, but if it occurs it will likely be retained.
Evolutionary theory in a nutshell.
Coming back to your claims about OOL research it seems clear that most scenarios may not seem to think that the first cell arose fully randomly.
Both SELEX and nature work very similarly and as has been shown both have the capacity for quite some innovation.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 09. January 2003 16:26
Frances:
You seem to be confusing randomness and unpredictability.
Nelson:
Where exactly do you see me confusing randomness with unpredictability? In this paper, Dermott is explaining why specific random events need to occur in order to get to the path to the functioning first cell.
You say:
quote:
Coming back to your claims about OOL research it seems clear that most scenarios may not seem to think that the first cell arose fully randomly.
In my post I mentioned why I think his paper is not irrelevant to natural selection. He focuses on specified functions. This focus is highly relevant to any form of SELEX experiments, which must first get specified function, in order to select it.
[ 09. January 2003, 16:28: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 10. January 2003 00:42
Nelson
quote:
Where exactly do you see me confusing randomness with unpredictability? In this paper, Dermott is explaining why specific random events need to occur in order to get to the path to the functioning first cell.
But he seems to have forgotten the selection part of it afaict. If the replicator arose much earlier than the first cell then surely one has to take into account these factors.
As far as Selex is concerned, it would work equally well if natural selection did the work. No requirement for teleology per se, just a selective pressure.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 12. January 2003 23:36
Francis:
But he seems to have forgotten the selection part of it afaict. If the replicator arose much earlier than the first cell then surely one has to take into account these factors.
Nelson:
I already address this above.
Francis:
As far as Selex is concerned, it would work equally well if natural selection did the work. No requirement for teleology per se, just a selective pressure.
Nelson:
In SELEX experiments they try to select for RNAs that perform a certain highly specified function. They take a random pool of RNA, followed by their selection based on a necessary interaction or function. This was in fact addressed in Dermott's paper. If it turns out that this process is highly improbable we may find the need for a teleological hypothesis at the OOL and in my opinion, IDers have this hypothesis locked in.
IP: Logged
|
|
|
Printer-friendly view of this topic