The Nirenberg and Matthaei experiment was a scientific experiment performed on May 15, 1961, by Marshall W. Nirenberg and his post doctoral fellow, Heinrich J. Matthaei. The experiment cracked the genetic code by using nucleic acid homopolymers to translate specific amino acids.
The Nirenberg and Leder experiment was a scientific experiment performed in 1964 by Marshall W. Nirenberg and Philip Leder. The experiment elucidated the triplet nature of the genetic code and allowed the remaining ambiguous codons in the genetic code to be deciphered.
The Marshall W. Nirenberg Papers Public Reactions to the Genetic Code, 1961-1968
Nevertheless, the problem of the genetic code at least in the restricted one-dimensional sense (the linear correlation of the nucleotide sequence of polynucleotides with that of the amino acid sequence of polypeptides) would appear to have been solved.
In the years after 1953, scientists scrambled to be the first to decipher the genetic code. In an attempt to make the race interesting, theoretical physicist and astronomer George Gamow came up with a plan. He organized an exclusive club, the “RNA Tie Club,” in which each member would put forward ideas as to how the nucleotide bases were translated into proteins in the body’s cells. His club had twenty hand-picked members, one for each amino acid, and each wore a tie marked with the symbol of that amino acid. The group—which did not include Marshall Nirenberg—met several times during the 1950s but did not manage to be the first to break the code.
Genetic memory resides in specific molecules of nucleic acid. The information is encoded in the form of a linear sequence of bases of 4 varieties that corresponds to sequences of 20 varieties of amino acids in protein. The translation from nucleic acid to protein proceeds in a sequential fashion according to a systematic code with relatively simple rules. Each unit of nucleic acid defines the species of molecule to be selected, its position relative to the previous molecule selected, and the time of the event relative to the previous event. The nucleic acid therefore functions both as a template for other molecules and as a biological clock. The information is encoded and decoded in the form of a one-dimensional string. The polypeptide translation product then folds upon itself in a specific manner predetermined by the amino acid sequence, forming a complex, three-dimensional protein.
Scads of scientists. Two Nobel Prizes. Isn’t consensus science grand?
“…the fact is that present life requires semiotic control by coded gene strings.”
– Howard H. Pattee
Well, yes. But someone chose that mapping, and it made sense for a specific reason. They originally had quite a different mapping for each letter and the story of how they chose the final one is quite interesting.
But nobody chose the genetic code, was really my point, for the reasons you say.
For others who may be interested, I’ve stated all along that the genetic code meets the mathematical definition of a code. I’ve also referenced coding theory. I’ll get to it all, in my own sweet time.
But nothing precludes anyone else from taking the initiative to look it up themselves.
As far as a single source laying it all out in glorious detail I can refer the readers to the following:
Information Theory, Evolution, and The Origin of Life
It’s a shame that even the kindle version costs so much. But as I’ve been saying, much of this can be found online. It’s not a mystery. No one is trying to hide this information.
Have you read The Information: A History, A Theory, A Flood?
I’m actually reading it right now. The pages have such nice edges.
That’s probably where I read the story, as it goes.
Can anyone answer? I posted definitions here.
I see a code as a two way transform.
I write “send the army at midnight”
It is “encoded” into “349jsdisjr934u08uefd” via some shared rule set.
It is “decoded” into “send tte army at midnight” but they get the idea anyway.
The genetic code is not a “code” like that because we cannot write “build an arm” and expect, well, an arm. The message cannot be “understood” as there is nothing to understand it. We have to send an instruction manual, step by step. And in the very first instance, chemical gradients are used to determine the starting point(s) for everything to “tell” each identical potential chemical reaction what to do. Some code that does not even know what page 1 is!
So the genome/DNA is not so much a code as a list of things that’ll happen from a certain starting point in a certain environment. It just so happens there are different reifications of certain aspects of it along the way.
my 2c.
The Real Genetic Code
Thanks, mung that’s helpful. As must be obvious, I know next to nothing about this stuff. The only books on evolution I’ve read have been by Butler–and he’s a Lamarckian! (I HAVE read four of those, though!)
But is this arbitrariness thought to be merely apparent–i.e., a function of our limited knowledge, or does it go “all the way down,” meaning that no amount of additional knowledge would allow us to successfully predict the pairings?
From what you’ve posted here, I can’t tell whether there’s actually a scientific consensus on this point. Have you read that Szathmáry chapter that’s cited?
Hi walto,
I’m going to try to channel Upright BiPed here. It goes all the way down. The actual mapping had to be discovered. The attempts to predict it failed. I think that is part of the reason the code is considered arbitrary.
Hi walto,
I don’t recall if I have read the book cited by Godfrey-Smith. But my memory sucks and if I have the book it’s in storage.
However, I do have here in my possession a copy of The Major Transitions in Evolution Revisited
I’ll muck about in that a bit and see if I can find anything.
Excellent question.
This is why origin-of-life theories are (currently) not well supported by evidence. Current thinking puts the very first life on Earth at about 3 billion years ago. For maybe 2 billion years organisms were, as far as we know, earlier versions of the bacteria and archaea that we still find today. No evidence survives of the biochemistry of these early organisms so the evolutionary pathways are a matter of informed guesswork.
That’s far from saying we can never work out plausible pathways or that we can never find additional evidence. Previously unknown micro-organisms still turn up in the unlikeliest places. We may even find something on Mars!
Mung,
No, I provded the technical details in support of my contention that the code is not a ‘true’ (ie representational) code. To handwave that away as, implicitly, obfuscation is to put no effort into understanding the argument. I try to avoid dumbing-down out of respect for readers’ intelligence.
The colloquial definition of ‘code’ is a system of rules for the conversion of one symbol set into another. The more rigorous mathematical definition refers again to source and target sets of symbols. My point all along has been that the code is not symbolic. The chain from input to output is causal, implemented in physics and chemistry. One could pedantically claim that Morse is physical too, but there is the interposition of an interpreted rule set which is absent in the genetic variety. An enzyme is not a rule.
One can map the code in a biochemistry textbook, as a matrix. It is very easy to turn the causal relationship into a symbolic one. But where are the rules? What makes a particular triplet ‘symbolic’? Within a particular cellular system a triplet a multiple of 3 away from the start of the reading frame causes a particular amino acid to be attached to a growing peptide chain via a series of steps. In every other circumstance the triplet is just 3 neighbouring bases. Either way, no-one has to interpret anything or look up a rule. There is just a population of molecules with amino acid on one end and anticodon on the other. Electrostatic interaction does the rest.
Fundamentally, codon-acid associations are determined by AARS enzymes and Watson-Crick binding of tRNA in the ribosome. If there were but a single AARS binding to one triplet, that would hardly be a code. But if there were two, suddenly there is something we could map. But it does not become a mapping in its physical operation simply because there are now two causal associations where previously there was one. We simply have an expanded population of tRNAs, and fewer STOPs.
That’s my argument, which you are free to take or leave as you prefer. Nothing hinges upon it; it’s just armchair semantics. Such a system can evolve; I don’t need to say ‘it’s not a code’ in order to pursue that. If ID depends so heavily on pushing people to accept definitional arguments, and moaning like stink when they don’t, it is on pretty shaky ground.
Mung,
Me too. You post long-known details as if they are hot off the presses.
walto,
There remain those (including, I believe, Larry Moran) who think that there is some kind of chemical relationship between a triplet and its acid. Personally, I’m doubtful; I think it really is chemically arbitrary. But that does mean it is ‘circumstantially arbitrary’ – that in the organisms in which it arose there was not some particular circumstance that channelled it down a particular path.
As I said in another post, although individual associations are arbitrary, the code matrix is not ‘random’. Acids with like properties group together.
My own view on the initial circumstances is that peptide synthesis was initially noncoding – that attachment of an amino acid to a short stretch of RNA (or even just an adenosine, a single base – helped to drive forward the energetically unfavourable peptide bond formation. This activation is part of the energetic landscape today; tRNA is not merely an inert ‘carrier’.
The attachment site remains a specific sequence in all tRNAs – ACC, adenosine and two cytosines. Coding arose later, when docking to mRNA commenced (this docking again helps drive the reaction – it is not merely an informatic association).
Triplet docking probably occurs due to the physical nature of RNA – four bases are hard to hold in a straight line on a looped structure and may bind too strongly for easy release; two would wobble too much and impart insufficient energy to the reaction. It is probably not the case that triplets were needed to allow for 20 acids, but for energetic reasons.
Very helpful, Allan. Thanks for answering Walto’s question to Mung (which likely would not have gotten answered otherwise, in my experience).
It seems to me that there might still be predictability at least in principle for why the mapping turned out the way it did, but not using the laws and models of biology. It would require the laws and models available from physics
Such predictability in principle involves assuming one can predict the history of the universe from the initial conditions. So it would have to assume that QM randomness was due to limitations in our knowledge only, and that limits on physical computation of predictions, such as chaos, are ignorable if one is talking about predictability in principle.
Moved a post to guano. Please remember that accusing another person of lying is against the rules. Assume, for the purposes of this discussion, that they are merely mistaken.
Or that possibly you are.
Tis the way of ID. Just be glad they are not underlining and bolding the “telic” words in a new OP at UD thereby “proving” ID is real. Or something.
Thanks for this, Mung. Can you channel the BiPed a little more, and tell us why you think this is important?
Allan Miller,
Does *not mean, I meant to say!
Do you mean that there might have been some other biological or perhaps chemical law which, given the pre-existing conditions, caused that particular mapping to be part of the genetic code?
Hey Mung: do you disagree with UB’s admissions regarding “semiotic theory?”
How would you answer my last question?
It has never been clear to me why an “arbitrary” relationship between the elements of a code presents a problem for natural selection. In fact, given that arbitrariness, codes can be “multiply realizable,” meaning that any number of mappings can perform the coded function interchangeably, just as thousands of human languages provide the functional advantages of any human language. Relative to an intricate determined mechanism that cannot broach an arbitrary element – and therefore allows just one (or many fewer) functional realizations – I would argue that the arbitrary element in coding makes it more likely that selection can hit upon and perfect an information-bearing adaptation, because many more variations in mapping would be adaptive, and functional rather than fatal.
Again considering the analogy with human language, which clearly has a strong evolutionary basis, were functional language necessarily non-arbitrary it would have been both much less adaptive (because inflexible) AND much more difficult for evolution by variation and selection to originate. Imagine if English was the only possible mapping between utterances and states of affairs. It would be vastly less likely that evolutionary processes would have hit upon not just the capacity for language generally (any one of thousands of infinitely flexible mappings will will do), but the capacity for English.
It is the arbitrariness of coded information transfer that places it within the reach of evolution.
BruceS,
I’d prefer ‘contingency’ to law. There are numerous elements of the system that may channel matters. I’ll try and summarise a few possibilities, taking as my basic (speculative) assumption an initial single non-protein AARS charging a single tRNA. There are two fundamental ‘ends’ of the association to consider: amino acid and anticodon
Acid end
– the set of available amino acids may have been restricted such that, though the triplet used in docking may be arbitrary, the acid thus ‘coded’ was not.
– the product of peptide polymerisation may constrain the association. If the cell has no use for polyalanine, but does have a use for polyguanine, this will channel initial associations towards using guanine.
Anticodon end:
– The production of a functional product through mRNA docking of tRNA depends on a particular series of triplets up to the first STOP (the first unassigned codon). As soon as an ’empty’ codon is hit, elongation stops. Therefore, the initial triplet which led to the first ‘useful’ peptide will have depended upon stochastic factors, relating to the pattern of random ‘code’ triplets and STOPs in the translated region. Attempted docking in useless regions would not be beneficial. It’s circumstantial, a frozen accident of sorts.
Subsequent amendment of the code takes place in light of what has already happened to it. Retooling an AARS to take a different acid and tRNA set would extend the acid library, but this is not unconstrained. There are two constraints on chemical property here:
– An AARS is more readily amended to take a similar acid than one with radically different chemical property.
– Substitution of this new assignment into existing proteins is less likely to be disruptive if chemical property is conserved.
Taking all this together, as a thought experiment I would expect the individual assignments to be completely different on other planets that had evolved a genetic code via the same mechanism, but I would expect all codes to be ‘fault tolerant’ in a similar manner to ours due to substitution constraint. Constrained arbitrariness!
Reciprocating Bill,
That’s probably the case, but it need not be a matter of actual history. Unless we assume convergence, all different codes would be descendants of some initial ancestor that started the basic system. There could conceivably be selection among these descendants so as to preserve those with ‘optimal’ characteristics, if survival were compromised by such relative imperfections, but I don’t think this is actually necessary to explain the ‘optimal’ features of the code – most notably, error correction. As I’ve tried to explain above, constraint on subdivisions of a simpler code inevitably gives error correction on individual misreads for free.
My expectation would be that all codes built in this way would be fault tolerant, so not much to select on among variants.
eta – this is not completely contrary to your point – amendment of a smaller acid set is possible precisely because associations are arbitrary, as you say.
very interesting discussion, thanks for elaborating so clearly.
If I interpret some of the ID posts here charitably, I would say it is not the arbitrariness that is their issue.
Instead, the issue is that the code is a language. Languages are more than just causation. Languages have meaning. Having meaninq requires that there are norms to distinguish right usage from wrong usage. I don’t meant syntax norms, which could result purely from biochemistry, but rather I mean semantic norms.
So in the case of the operation of the transcription mechanism, some transcriptions fail to preserve the meaning of the code (that is, the process breaks down somehow and the correct protein according to the code does not result). But why that is mistake cannot be attributed to causation, since the breakdown itself would have a causal explanation. Why is one causal process right and one wrong?
So if the genetic code has associated norms, where did they come from? How can a purely causal process create norms?
I think a teleosemantic explanation is easy to make work here. We combine the causal process of development that Allan has outlined with the teleosemantic explanation that norms originate by asking what was selected for during that process to promote fitness. In some cases, due to physio-chemical laws, there may have been only one choice to select from. In others, there may have been contingency and channelling, which I understand Allan to use to refer to a form of selection (eg as in his anticodon example and perhaps in promoting optimality, though perhaps not in this case).
Another approach, which I understand KN to prefer, would be to define life as having a purposes, which at their core would include survival and reproduction, and then derive norms according to what best promotes that. Then a right transcription would be defined in terms of what best enables the organism to continue to live, where “best” can now be defined purely in causal terms.
To the surprise of absolutly no one, Brave Sir Mung has waddled back to UD and, er, mistakenly commented about what has gone on here.
Of course multiple people answered the question on that thread but poor Mung just can’t bring himself to pass the truth to the IDiots.
OK, this is really interesting stuff; much appreciated. But I think if you put it together, you get a couple of the premises Frankie has insisted on (at least in my reconstructed argument–which may be completely off base).
Let me put it again:
Allan concedes the likelihood of (2) in his view. DNA connections really seem to be arbitrary–not in being currently unpredictable because of our current ignorance, but because they are in some relevant sense “outside rules”–random.
And RB seems to endorse part of (4) here. He refers to multiple realizability, languages, etc. Stuff that we generally see connected only with mental activity. This gives Frankie a good chunk of what he/she wants.
(1) is hard to deny, I think, but, assuming Allan and RB are both correct there remains a bit of the inference to (5) that would need bolstering. And it’s likely the question-begging part of the argument, which is analogical. Viz., the move from “OK, maybe DNA sequencing exhibit an arbitrariness in just the way that mental activities and in our experience ONLY behaviors at least partially resulting from mental activities also exhibit” to “It makes sense to infer from that unique similarity that the DNA sequencing also results from some mental activity.”
As it remains no more than an argument from analogy, it can obviously be attacked at that point. But I’d have thought that a number of scientists might resist (1), (2), and/or (3) as well, so that the argument would never have gotten all the way to that point.
Again, many thanks for the elucidating comments.
Just wanted to add that I find it curious that religious arguments have, at least since Aquinas depended on purpose-not-randomness; that was Paley’s whole bag of tricks: “How (he bellowed) can we get something as exquisite as the human eye from random physical processes?!”
But here, it’s precisely the existence of randomness that is depended on to prove Design. “Purely physical processes are orderly, never arbitrary. But DNA sequences?–Why, they’re like human languages, the connections are wholly random. Lots of other choices could have been made.”
As I said much earlier in the thread, I’d never heard this argument from arbitrariness before. And when you consider the history of Design arguments, it’s really kind of amazing.
walto,
‘A particular sort of predictability’ must include the predictability of stochastic systems. And that’s what we are dealing with here. There is no fundamental necessity that one of Lenski’s 12 lineages would evolve citrate uptake. One did, 11 didn’t. But there is a regularity going on here: presumably unguided exploration of nearby genetic ‘space’. You can’t predict in the sense that is frequently used, any more than you can predict the final resting place of a dropped stone. But you surely don’t need angels to lay it down.
It displays many regularities, both in faithful and erroneous modes. You just can’t say what’s going to happen next.
Would this also apply to the weather? I don’t see how this helps ID either way.
I believe the weather prediction issues are a function of ignorance: if we really knew all the pre-existing conditions (barring any quantum randomness) we could predict weather with perfect accuracy. I took you to say that ignorance isn’t the problem with DNA predictability.
As to how it helps–as indicated, it firms up the premises they need to make an analogical argument. That’s all it CAN do.
Just skimmed Torley’s piece over at UD.
If your first piece of evidence is a non-peer-reviewed letter, and an appeal to authority to boot … ! It should be noted that Crick was not an expert on semantics or linguistics ***, great scientist though he was. He rued his choice of the word ‘Dogma’ for example.
*** no more am I!
Torley loves the appeal to authority. Jonathan Wells has two (count’em!) PhD’s.
That is evidence only that the term has come into accepted parlance. The definition of ‘code’ has even been extended to accommodate that sense. I’m happy to use it. That does not mean I think it a code in the same sense as Morse or ASCII. It can be rendered as such, but is not itself.
His third piece of evidence is actually his first, but used by someone else – Meyer quotes Crick.
His fourth is Jonathan ‘Two PhD’s Wells.
So that’s Crick, the common usage of the term ‘genetic code’ by scientists everywhere (including me), and two ID proponents. Proving what? That people are happy to call it a code? That was never in dispute.
But it is not symbolic. Can you change the ‘font’ of a triplet and end up with the same informatic result? Does the symbol UUU represent phenylalanine? No. At best it ‘represents’ tri-uridine. The physical trimer represented by ‘UUU’ does not represent phenylalanine either. In some circumstances it causes phenylalanine to be attached to a peptide (indirectly). That is not a representational relationship.
Its code-like properties are bestowed upon it by us.
Multiple realizability is not a peculiarity of minds. We see in it many systems due to selection effects. Examples
– flying between birds, bats, insects
– pain between octopuses and us
– territorial behavior among many different species.
There is a further, different discussion about whether the details of the human mind really are multiply realizable, or whether our biochemical implementation is the only way to duplicate the detailed relationship to the body, and the timings constraints inherent in using our bodies in the world. That is very much an open question. I suspect the answer is that the human mind in all of its detail is not multiply realizable.
I see Allan has already made points I was also thinking about in your (1) and (2). Even QM, our best prediction, includes randomness. There is nothing unscientific about it being included it in understanding the origins of the particular genetic code we see.
Only partly true.
Sunspot activity, forest fire smoke, volcanic ash in the atmosphere can all affect the weather. These are all outside typical weather models and would be considered stochastic effects by those models.
That does not mean those outside factors do not have a deterministic basis in physics, at least in principle.
I think the same is true of selection effects.
walto,
It is in the same sense that the weather is, but I’m not too sure about invoking hidden variables or perfect knowledge. You just treat as ‘random’ that part of the picture you can’t get at, regardless whether it is even get-at-able in principle or not. In this particular instance, it’s in the past, so that’s even less accessible.
We just don’t know where a particular mutation is going to occur, nor what its relative benefit will be. These are circumstantial. Some mutations are more likely than others depending on the molecules at and around that site; torsional strains and mechanical damage, etc etc etc. Within the population some beneficial mutations are lost, some detrimental ones preserved. But there remains an arrow, a bias in the direction of improved relative fecundity.
If the same IS true of selection effects, then there isn’t arbitrariness all the way down. I take no position on that–I’m following Allan here.
Re multiple realizability, I just threw that in. You’re right that it was a bit of handwaving on my part.
This is the key issue for premises 1 and 2, I think. The speculation about “perfect knowledge” is required to determine whether the apparent arbitrariness is real or simply a function of our ignorance when making predictions at some given time. If that’s all it is, then nothing follows. I took you to be claiming (albeit tentatively) that the arbitrariness in these DNA connections is not of a merely epistemic kind, but exemplifies a real absence of determination.
This ties in to Bruce’s comment about sun spots as well. The argument isn’t about defects in some current model; it’s a claim about real non-determination, regardless of the perfection of the model. I think it needs that (though I probably should think about its effect on premise (1) a bit more).
You might “treat as random” that part of the picture that you can’t get at. But that approach is not a logical necessity but is simply a choice on your part.
I would say that there is no way for a subjective observer inside the universe to ever determine if the apparent arbitrariness is real or not.
This is just the problem of other minds.
peace
I’m at a 4 1/2 hour performance of Tannhauser (“Met Live in HD” – it’s intermission). Back later – MUCH later.
Wow. I love the overture, incidentally.
Maybe, but In the analogous area of quantum theory, we see a lot of a type of disagreement on this issue that suggests that participants on both sides seem to think both that there’s a fact of the matter and that it’s knowable. Similarly, hard determinists seem pretty sure that there’s no (real) arbitrariness in the world.
I’m not saying you’re wrong, just that I don’t think the claim that it’s impossible to know one way or the other (unless we find a rule that disconfirms the arbitrariness theory) isn’t trivial.
I would agree that there is a fact of the matter and that it is knowable. I just don’t think it’s knowable sans (wait for it) revelation.
Funny how that works isn’t it
peace
Well, then, which is it?–The arbitrariness “Frankie’s argument” supposes to show the existence of Design by noting similarites to cogitation in some apparently purely physical biological processes, or the non-arbitrariness that Calvinism requires?
ETA: Actually that’s not really fair of me: I put a false dichotomy. I think Calvinistic determination is really a type of fatalism supposed (incorrectly, IMO) to result from God’s foreknowledge; it’s not a scientific determinism.
When every question can be answered with the same answer, that answer is worthless.
Correct except for the fatalism part.
A Calvinist can hold that there is real arbitrariness in the universe. A materialistic determinist simply does not have this option.
check it out
from here: http://dictionary.reference.com/browse/arbitrary?s=t
quote:
Arbitrary: subject to individual will or judgment without restriction; contingent solely upon one’s discretion:
end quote
Maybe the existence of free will is not a red herring in this discussion after all.
peace
fifthmonarchyman,
It’s not a choice if I can’t bleedin’ get at it!
It’s your choice to perform the actions that prevent god from revealing those truths to you though, init?
I think we’re drifting somewhat. When we say a codon assignment is arbitrary we merely mean that there is not a causal relationship between the chemical properties of a given triplet and the chemical properties of the amino acid it … er … causes to be linked into a peptide.
Of course, once we have one codon assignment, the next is not-quite-so-arbitrary, for several reasons. But again it is not dictated by properties of the codon or the acid.
The interesting thing on causality here is that a single, tiny change in a tiny corner of a cell somewhere has effects that ripple through the ages, due to selection/drift. In the case of the fixed codon assignments, that’s every living thing, everywhere. There is certainly no way to predict what tiny corner-of-a-cell change is next.
OMagain,
If I knew what I was doing wrong … !