A computer programmer noticed that he was not able to type very much in a single day. But he mused that if there were a large number of software bots working on his code, then they might be able to proceed via totally blind trial and error. So he decided to try an experiment.
In the initial version of his experiment, he established the following process.
1. The software was reproduced by an imperfect method of replication, such that it was possible for random copying errors to sometimes occur. This was used to create new generations of the software with variations.
2. The new instances of the software were subjected to a rigorous test suite to determine which copies of the software performed the best. The worst performers were weeded out, and the process was repeated by replicating the best performers.
The initial results were dismal. The programmer noticed that changes to a working module tended to quickly impair function, since the software lost the existing function long before it gained any new function. So, the programmer added another aspect to his system — duplication.
3. Rather than have the code’s only copy of a function be jeopardized by the random changes, he made copies of the content from functional modules and added these duplicated copies to other parts of the code. In order to not immediately impair function due to the inserted new code, the programmer decided to try placing the duplicates within comments in the software. (Perhaps later, the transformed duplicates with changes might be applied to serve new purposes.)
Since the software was not depending on the duplicates for its current functioning, this made the duplicates completely free to mutate due to the random copying errors without causing the program to fail the selection process. Changes to the duplicated code could not harm the functionality of the software and thereby cause that version to be eliminated. Thus, in this revised approach with duplicates, the mutations to the duplicated code were neutral with regard to the selection process.
The programmer dubbed this version of his system N.E.C.R.O. (Neutral Errors in Copying, Randomly Occurring). He realized that even with these changes, his system would not yet fulfill his hopes. Nevertheless, he looked upon it as another step of exploration. In that respect it was worthwhile and more revealing than he had anticipated, leading the programmer to several observations as he reflected on the nature of its behavior.
Under these conditions of freedom to change without being selected out for loss or impairment of current function, what should we expect to happen to the duplicated code sequences over time and over many generations of copying?
And why?
[p.s. Sincere thanks to real computer programmer OMagain for providing the original seed of the idea for this tale, which serves as a context for the questions about Neutral Errors in Copying, Randomly Occurring.]
And our point all along has been why do you think it’s a problem?
If the sequences don’t currently have a function then they are effectively randomized anyway. The probability of an additional mutation causing any particular sequence to produce a novel function doesn’t change regardless of the history of the sequence.
You’re still making the beginner’s mistake of assuming evolution was somehow aiming for the specific genetic sequences we see now and that only those sequences will work for viable life. The number of simple biological concepts you’ve been unable to grasp on this thread just make me shake my head.
thorton:
mike elzinga:
Like Frick and Frack.
thorton, not having read Meyer’s book, comments.
Mike, having not read Meyer’s book, comments.
“The Skeptical Zone”
Good job Lizzie!
thorton:
How so?
Let’s say, for the sake of argument, I give you a completely novel sequence having no similarity to any sequence from a living thing. It has exactly one base pair “broken” so it does not code for a protein.
How do you distinguish it from a random sequence?
petrushka:
If I understand you, it has “no similarity to any sequence from a living thing” except for the fact that “it has exactly one base pair ‘broken’ so it does not code for a protein.”
Skeptical much?
Evasion and lack of answer from Truth Lover Mung noted. Again.
That would be the wrong conclusion to draw, and also not the point of my argument concerning Hoffmann. For starters, Hoffmann himself believes it is answerable (though he cannot say if we will find the answer in a thousand years). Secondly, its not just Hoffman that’s stumped. No one has a good answer.
A main point of highlighted Hoffmann’s statements is this. There are those who operate from the dismissive assumption that if someone such as myself thinks that no one has a good answer about how evolution could produce the complex molecular machines and systems we find in cells, then it must mean that they don’t understand the basics of evolution. To show the error of that assumption, all one needs to do is take me out of that picture and substitute Hoffman.
Does Hoffman believe in evolution? Yes.
Does Hoffman have at least a basic understanding of the principles of evolution? Yes.
Does Hoffman understand and believe that “complex biological systems can evolve from simpler precursors, and that the precursors don’t need to be as effective or even have the same function as the extant system”? Yes.
Understanding all of this, does Hoffman therefore think that anyone — not just himself, but anyone at all, including the most adept proponents of evolution — has a good answer for how these processes could produce the original molecular machines? No.
He certainly thinks it has something to do with replication and hitting on a molecular machine. There is a conjecture related to deriving from enzymes. But none of that changes his clear statements about the fact that no one knows how to apply these evolutionary processes in a way that would actually get us to the original molecular machines.
The relevant conclusion: The problem is not one of whether someone understands evolution or not. It is about the inability (so far, at least) of anyone to apply that understanding in a way that plausibly gets to the original molecular machines.
Corollary: It’s nothing against you that your case doesn’t get there either.
Notice one particular example of this. Here is what you repeatedly keep coming back to…
Yes, you’ve mentioned the peptide bond more than once or twice now. But, despite claiming to specify advantages for every “step”, the next specific selectable advantage you provide is not until after the TRS is operational. That is clearly a long, long way from your peptide bond advantage. However much you claim to only be taking “small” steps, that is a grand leap, not a small step. I realize you believe that the gap is filled with many small steps, but your claim was to have specified advantages for every step. The distance between these advantages that bracket the TRS is huge.
While you come back to the peptide bond repeatedly, what I’ve repeatedly asked you for is the manner in which evolutionary processes can plausibly be expected to implement the algorithm of sequentially reading a sequence as triplets.
You allude to having explained this elsewhere. Perhaps you could at least begin by summarizing the key points? How does a system that gets none of its reproductive advantages from reading a sequence as triplets find its way to building such a process?
What Mung doesn’t understand is that most of us knew before Meyers’s book came out what was going to be in it. It was merely a matter of watching the specialists in this area review it and confirm it. Once that was done, there was no need to read it.
It is very similar to the way we can know that Mung will always hurl feces but never address anything of substance. Just as we know he will issue a tu quoque on every observation of his predictable behavior. He does it every time; it’s all he knows how to do. No need to review it; it’s just that simple.
Mike Elzinga:
You and all the other alleged skeptics here at TSZ, including Elizabeth. So why read it.
Not a skeptic. Lizzie would be proud,
Wrong. We think you don’t understand the basics of evolution because you’ve demonstrated that you don’t understand the basics of evolution. Like not knowing mutations can’t be labeled “deleterious, neutral, or beneficial” until their effect in the current environment is known. Like offering Sanford’s “genetic entropy” as a valid scientific claim. Like not understanding things such as exaptation and that biological systems and their parts can take on different functions over time. Like thinking “we don’t know yet” means “it’s impossible.” That’s just to name a few.
Tell us Truth Lover Mung: what new ideas did Meyer put forth in Darwin’s Doubt that weren’t already presented in the film Darwin’s Dilemma and in his other various propaganda pieces on the web? Every review on the book from every paleontology expert I’ve seen points out it’s just the same old crap from Meyer, including his usual whine “evolution can’t explain all the new biological information!!”
Why are you unable to discuss Meyer’s ideas?
thorton
Please create an OP on “the basics of evolution.” And explain “the basics of evolution.” Unless you don’t understand “the basics of evolution.”
thorton:
I’ve read the book, you haven’t. Why are you unable to discuss Meyer’s ideas? Because you haven’t read the book.
Skeptical much?
I’m perfectly willing and able to discuss Meyer’s ideas. You seem to be the one who always finds something else to do when the questions show up.
Here’s a summary Meyer provided today at EN&V:
Meyer: “Instead of exemplifying a fallacious form of argument in which design is inferred solely from a negative premise, the argument for intelligent design formulated in Darwin’s Doubt takes the following form:
Premise One: Despite a thorough search and evaluation, no materialistic causes or evolutionary mechanisms have demonstrated the power to produce large amounts of specified or functional information (or integrated circuitry).
Premise Two: Intelligent causes have demonstrated the power to produce large amounts of specified/functional information (and integrated circuitry).
Conclusion: Intelligent design constitutes the best, most causally adequate, explanation for the specified/functional information (and circuitry) that was necessary to produce the Cambrian animals”
Premise one is complete bullshit with the usual IDiot undefined terms and unwarranted assumptions. That makes the conclusion worthless.
You willing to discuss the idea? Or do you have to go wash your hair again?
thorton:
Sure you are. Let us know when you’ve read Darwin’s Doubt.
I read Meyer’s direct description of Darwin’s Doubt he published today at EN&V. Posted it right above. I even cross-posted it to the Meyer thread so you’d have two places to run wimpering from any discussion of the ideas in DD.
Your “BUY THE BOOK” excuse wore thin some time ago Mr. Truth Lover. Time for you to find a new excuse for evading discussion.
ericB,
Eric:
On Hoffmann. Yes, I think he has a good grasp of evolution. But he’s not an evolutionary biologist, nor a molecular biologist, nor a chemist, nor a geneticist. He may or may not be aware of or accept the RNA World hypothesis. Now, if I was talking to Hoffmann, I might discuss these things, and perhaps he would agree that there is a germ of plausibility in my schema . Perhaps not. But, though you have explained repeatedly, I don’t see how substituting Hoffmann’s slightly hesitant comments in that Q&A for you gets us anywhere. You could find any number of people who, with slightly more relevance , are dismissive of the RNA World hypothesis. I could find any number who think it correct. In either case: so what?
This is precisely why I want you to focus. I know you think I’m dodging, but you have seen my proposal twice at least, in outline, and leap about the sequence like a man possessed, saying this-can’t-happen and that-can’t-happen. Others have accused you of Gish Galloping, and I can see their point. Before we go any further, do you accept the proposal that an organism with transcription but no peptide synthesis could plausibly evolve ribozymes that can make aminoacyl-AMP, can use that to aminoacylate a short RNA, and that such RNAs could plausibly perform the chemical feat of generating a peptide bond, and that the product at each step could plausibly possess selective advantage for the organism that produced it, purely from a process of random variation (mutation, recombination and drift) and selection? If the answer is ‘NO’, why waste each other’s time further?
poly-G codes for a polyglycine anchor sufficient to attach the RNA to a lipid-aqueous interface. Gly3 to Gly5 would do. And the probability of finding Guanine15 is not 1 in 4^15.
Maybe CCC coded for Alanine back then. Maybe cytosine9 was all you needed.
I have not said anything that would suggest that that was my supposition.
Err, “Observation #3: For any sequence that is not providing a selectable reproductive advantage, random neutral errors in copying will accumulate randomly throughout the sequence.”
While individual VSDMs will occasionally fix, they don’t accumulate because the back-mutation fixes more often. According to your reference, Akashi,1995.
The real problem is with the incoherence of “any sequence that is not providing a selectable reproductive advantage”, but you seem unable to understand that point.
Eric,
I’m not denying that there are gaps in my suggestion. There are always going to be gaps, because however much detail I supply, you can demand that I go to a finer scale. Ideally, one would have every single base change in every organism along the way, and a causal account of the biochemical and selective milieu that its possessors encountered (necessarily requiring access to organisms that the process, by its very nature, has rendered extinct or radically altered). You know that’s not going to happen, just as I know you are not going to supply any detail whatsoever for your alternative account of Design. All anyone can do is apply a fairly broad brush. But those initial three stages – aminoacyl-AMP, aminoacyl-ACCNNN(…), and peptidyl transferase using two or more of the latter as substrate – these form a useful test bed for concepts.
If you don’t think any of those steps can occur, then you don’t really think evolution can occur.
Or perhaps the absence of precision in my proposed selective benefit forces it to remain moot. It would be nice if I could give THE reason for the selective benefit of aminoacyl-AMP for example. I can only point rather vaguely to it being an initial step in certain metabolic reactions. It transfers ATP energy to the aminoacyl-AMP complex, making it more reactive.
But it is also possible that the two-step process,
aminoacid + ATP -> aminoacylAMP
aminoacyAMP + ACCNNN(…) -> aminoacylACCNNN(…) is actually the one that produces selectable benefit. Such a benefit could be transport, or localisation in a particular part of the cell due to the chemistry of the ACC ‘tail’.
So how does a 2-step process arise if the first does not provide a benefit? Well, for one thing, it’s chemistry. Enzymes don’t generally make things happen that wouldn’t happen at all otherwise, they increase yields and lower activation energies for processes that are entropically favoured anyway (albeit only with energetic coupling to NTP in many cases). There is a nonenzymatic citric acid cycle, for example; the enzymatic version smartens the whole thing up and allows for more subtle control, but it’s not doing the chemically impossible.
And so with a 2-step reaction. Aminoacylation of an ACC stem could conceivably proceed nonenzymatically, at low yield. The reaction could be 2-step, via an aminoacyl-AMP intermediate. A ribozyme could arise that catalyses the first (or the second) step, which would drive the equilibrium towards the product. A second ribozyme could arise that catalyses the other, increasing yields still further and allowing for finer control.
Another way this could happen is by drift and recombination. By drift, a non-deleterious sequence can become resident in a population at a nonzero frequency. Although the A+B type can only arise from either the A or the B subpopulations by mutation, the probability is substantially enhanced by recombination between members of the A-only and B-only subpopulations.
You will notice, and leap on, extensive use of the word ‘could’ in the above. Your demand for causal detail is understandable, but the nature of the processes does not lend itself to absolute precision in reconstructing history. The possibilities for evolutionary amendment are far greater than your assumptions would allow, and randomness plays a big part (take note of Hoffmann’s statements on the importance of randomness for novelty). This is part of the problem. The possibilities are enormous in detail. And it’s not solely a story of iterated incremental change + deterministic selective benefit or loss, step by tiny step, one thing at a time and ever upwards.
With these points in mind, if you don’t accept as plausible the rationale for these first steps, we aren’t ever going to get there.
Incidentally, I didn’t crib my sketch from this, but I could easily have!
Origin and evolution of the ribosome
A couple of variant ideas for the original role of mRNA are offered. One is that tRNA docking helps stabilise the tRNAs to increase the likelihood of reaction instead of Brownian drift – though ISTR that docking itself provides some of the energy for the peptide bond, improving the thermodynamic profile. These aren’t mutually exclusive. Another suggested function of the not-yet-mRNA in the paper is the delivery of tRNAs to the Peptidyl Transferase Centre.
As to ‘why triplets’ – well, that’s the size of an anticodon, formed as it is from folded RNA that base pairs with a more relaxed single strand. Not as facetious an answer as it may sound. Could it be smaller, given the material constraints and the putative role? I don’t think so. Could it be larger? Slightly more likely. But whatever size it was, we’d be wondering why, and though I think there is a case for optimality, another possible answer is ‘because that’s how big the first one was’.
thorton still thinks that “read the book” means “buy the book.”
Mung’s complete avoidance of all attempts to get him to discuss the ideas in Darwin’s Doubt as expressly described by Meyer himself is noted.
Again.
I could perhaps be enticed to discuss with thorton the ideas in Darwin’s Doubt as expressly described by Meyer in that book, but alas, there’s no evidence at all on record to indicate that thorton has read Darwin’s Doubt and sufficient reason to believe that he has not
Mung’s complete avoidance of all attempts to get him to discuss the ideas in Darwin’s Doubt as expressly described by Meyer himself is noted.
Yet again.
Mung loves The Truth don’t you know. It’s just defending Meyer’s claims that he hates.
thorton, not having read Darwin’s Doubt, flounders in a sea of incoherence.
Had he read the book, he could have discussed it’s claims with me. I even started a thread on the book for that express purpose. But thorton is a true skeptic. Lizzie would be proud.
Mung’s complete avoidance of all attempts to get him to discuss the ideas in Darwin’s Doubt as expressly described by Meyer himself is noted.
One more time.
Mung making excuses and avoiding all discussion. You can set your watch by it.
With an important deadline approaching (as well as other demands), for me that has tended to push responding here to the back burner lately, so posts may be less frequent at present.
“leap about the sequence”?? I began to focus specifically on the part of your outline for The Triplet-Reading System on September 7, 2013 at 10:42 pm. That was back in the original thread, more than 6 weeks ago. Since that time, I concentrated on a series of posts related to the theme of the TRS, often referencing The Triplet-Reading System explicitly in a heading.
This current thread is itself an expansion of a point made in the former with the TRS in mind. For more than 6 weeks, out of your whole outline, its been the main illustrating example under consideration, pretty much exclusively. (A main point is the problem of the implausibility of success via a blind random walk of any significant distance from selectable advantageous features. Hence the constraining requirement to find a chain of sufficiently neighboring selectable stepping stones.)
And in all those 6+ weeks, the main point I’ve found most difficult to arrive at is plausibility for an evolutionary path that leads to implementing an algorithm to process an RNA strand iteratively and sequentially, i.e. how we expect to actually get to something that discovers reading a sequence of triplets sequentially — a fairly central feature of any Triplet-Reading System, wouldn’t you agree? I started focusing on the iterative algorithmic nature of the TRS way back on September 12, 2013 at 1:19 pm in The Triplet-Reading System — Part Two.. That is just a couple days short of 6 weeks ago.
If you had accused me of being too focused on just one part of your outline, I would have at least seen your point.
Yet, the TRS has served well to illustrate the difficulty of dealing with certain actual hard parts — the parts that you realize I think you are largely dodging and/or dealing with in a more hand waving manner. I don’t find it at all surprising that you major on the particular points you emphasize, while keeping certain other aspects in a vaguer background (or even hold them hostage unless and until I affirm the parts you prefer to focus on).
thorton:
If thorton has read the book and wants to discuss it there is a thread available to him created for that express purpose
Mung’s complete avoidance of all attempts to get him to discuss the ideas in Darwin’s Doubt as expressly described by Meyer himself is noted.
One more time, again.
Pity that all Lover of Truth has is making excuses and avoiding discussion. But there it is.
ericB,
Eric,
I accused you of ‘leaping about the sequence’ because you keep ignoring these basic principles and jumping ahead to the ‘triplet’ part. Because the early steps make no reference to triplets, the appelation TRS is unhelpful. You grumble because my explanation of a particular step is not an explanation of a TRS, but of something that isn’t a TRS until the day it starts docking carriers against an RNA strand. And you spin off into avenues of misunderstanding of neutrality and selection, and a reliance on analogy instead of direct consideration of the phenomena in question, which do not advance discussion.
If you think I’m dodging, ask me a specific question and I will attempt to answer it. But if you want to know ‘why triplets’, you need to have some appreciation of what I think went before, and why your ‘NECRO’ model and cave fish are not relevant.
But if you continually dismiss the groundwork – prior steps which were clearly not ‘triplet reading systems’, but which form the basis for amendments that ultimately led to a system that is – then conversation is difficult, because it appears you are simply out to deny, not evaluate. “Bring me your rationale, and I will deny it. Whatever it is”. That, at least, is how it appears from this side of the computer screen.
But there’s no point in devolving into a meta-argument about the conversation. With a clean slate: why do you think uncoded peptide synthesis is not admissible as a rational step, beneficial in its own right without any regard to where subsequent amendments would take the system?
And … I have addressed the ‘vague and hand-wavy thing’ at great length. I do not know the categorical benefit of anything that happened in organisms of unknown constitution in an unknown environment 3.5 billion years ago. Things are ‘vague and handwavy’ not least because there are actually several possibilities, and no means to adjudicate between them.
I will here trot out the now-familiar observation that you haven’t the faintest idea what the Designer did, or when, so if you have a problem with the vague and handwavy, you might care to be a little more careful when dishing out critiques.
I can offer plausible benefits, and a sequence that accords with techniques available for elucidating such things. But one can never ‘prove’ history. So if you think that this lack of absolute definitive Proof proves Design, and that satisfies you intellectually or emotionally, then I have nothing to say that would make you think otherwise.
ericB,
Indeed. As I said, I was in Norway from then until 30th September (I had a great time, thanks for asking!), and had not been to TSZ for several days before that, nor several days after. There was a great deal to catch up on, too much to sensibly address.
One repeated theme appears to be a fundamental disconnect between your viewing of the process as an algorithmic one, and mine as a chemical one. I’m a biochemist by training but a programmer by trade, so I do have a foot in both camps. I would advise against thinking too much in algorithmic terms, myself. I think it’s misleading. I would also require some better evidence that ‘function’ is islanded, in the kinds of system we are talking of. The space is not fully connected, but nor is it definitively discretely lumped. You simply assert this.
You make the assertion that “the probability of finding Guanine15 is not 1 in 4^15” but without providing your basis for the claim. Perhaps you would be willing to provide that?
There are no chemical bonds directly between the bases, so there is no direct chemical requirement for any particular sequence of nucleotides to be preferred over another. That is exactly why DNA and RNA are well suited as flexible carriers of information, since they are free to form any sequence without preferential bias. Therefore, a claim of having other than random probability would need to be justified by something other than the chemical requirements of DNA or RNA.
In addition, it would not be legitimate to employ any implicitly teleological assumptions, such as just assuming that in some unspecified way there is a predisposition to produce and preferentially use lengthy repetitions of the codon pattern that just happens to match the first (and only) tRNA. I mentioned the issue of unwarranted teleological assumptions regarding sequence back on Sept. 14.
I expanded on the issue in a p.s. the same day. The problem is not just overcoming the improbability of consistently producing working input sequences that match the one anticodon (so as to produce a detectably significant benefit that is greater than noise variation). To be plausible, the system has to be able to develop to have additional tRNA that match different codons. That means producing and providing other sequences that work with the 2nd tRNA, and then the 3rd tRNA — all without making teleological assumptions about what the system could know or prefer.
Random sequence input is problematically improbable. But a constrained sequence ordering presents both problems of justifying the constraint as well as problems for transitioning to a more elaborate set of tRNA that requires violating any such constraint — yet without falling back into the improbability of random sequencing.
Many people have long been making the claim that much of the genome is just junk. Even if that turns out not to be true, that wouldn’t make the concept incoherent. Why would you think the idea of a “sequence that is not providing a selectable reproductive advantage” is incoherent?
BTW, when you state, “While individual VSDMs will occasionally fix, they don’t accumulate because the back-mutation fixes more often.”, that does not affect my point. The only way that a back-mutation can fix is if that is referring to a case where there is some selectable reproductive advantage to be gained by the back-mutation. (Otherwise, a back-mutation would be equally neutral and not preferred compared to any other copying error.) If so, that is not the case I have been discussing.
Throughout the thread, my Observations have been focused on those cases that are sufficiently neutral that there is no significant loss of reproductive benefit due to the copying error. Consequently, such copying errors must accumulate randomly whenever those conditions apply. (Throughout the thread, I’ve also always acknowledged the obvious fact that there are other situations where this is not the case, and I included this explicitly in my Observations.)
You did make the point early on that even a slightly deleterious mutation might not be weeded out by selection but be lost in the noise. When that is so, well and good. Your point has merit. When that is not the case, i.e. when selection can weed out the change by preferring back-mutations, fine — that is a different case. But it has no effect on the reality of the existence of changes that are not weeded out, particularly when the sequence is providing no reproductive benefit and therefore no basis for preferring any back-mutation.