Have you ever tried writing palindromes? How about writing phrases that can be read the same way in either direction? Here are some examples:
A man, a plan, a canal: Panama
Live not on evil
Was it a car or a cat I saw
These sentences were no doubt designed…
Can you imagine writing a book that can be read forwards and backwards containing 2 different stories that made sense? Not an easy task…
Watch the video and pay special attention to the following examples:
- Alternative splicing of RNA that produces multiple proteins from one gene
- Duons – Overlapping sequences that code for both protein expression and transcription factor binding sites simultaneously
- Dual coding genes in which one sequence is read in multiple frames to produce completely different protein
The magnitude of the dual coding problem in DNA would be the equivalent of writing a novel that could be read in either forward over reverse directions making two different stories both of which made sense…
And don’t forget that according to Darwinists the dual coding in DNA simply evolved, right?
No, that’s not my claim. Look again. Please try harder to make sense, both when reading and when writing.
Dunno. What papers are you citing here?
Are you really claiming that DNA repair is restricted to pre embryo development? Can you support this claim?
No, you apparently don’t understand the claim. Of course DNA repair happens all the time in all cells. But that repair happens soon after the mutation or lesion, or it doesn’t happen at all. As has been explained, a point mutation can only be repaired (back to its prior condition) if there is a way for the repair enzymes to tell that it’s a mutation, and the only way to do that is if it’s changed to a non-standard base (uracil, e.g.) or if there’s a base mismatch with the opposite strand. In the latter case, there’s no way for the enzymes to tell which base is the original. And once the strand is “repaired” to the mutation, that’s the end of repair.
Entropy,
In order to play around to any degree beyond a few mutations you need a huge amount of functional space or when you start to play around unless through serendipity you find a beneficial mutation in the initial mutational change ( malaria resistance as an example) the sequence will rapidly move to non function. There may be proteins where playing around is possible but they don’t live inside the cell nucleus where mutational sensitivity is high and binding to multiple protein types is required.
This is why I think the Darwinian mechanism is losing steam as a plausible explanation for life’s diversity.
Actually the only thing that is necessary is that there are other potential functions in the immediate neighborhood in that “space”. There doensn’t really have to be a “huge amount”.
And actual experiments have been done on this to try to figure out whether there is “something functional nearby”. See for example: Can an arbitrary sequence evolve towards acquiring a biological function?
Unless there is purifying selection.
[citation needed]
No wonder you do, all your premises are wrong.
Has anybody tried to apply Joe Felsensteins’s formula in the improving palindromes with the amazing creative abilities of the destructive, random mutations?
The sheer dumb luck has gotta be more intelligent than a couple of Phd’s? At least that’s the way it appears to be… I personally think that all creative work should be left for sheer dumb luck to do it right… Why bother spending billions of dollars on research that can’t figure out how sheer dumb luck did it.. Sheer dumb luck of the gaps…
Rumraket,
This is consistent with my claim.
Rumraket,
It’s the exact opposite of your ignorance based claim.
I can find nothing in this paper which substantiates your claim that “There may be proteins where playing around is possible but they don’t live inside the cell nucleus where mutational sensitivity is high and binding to multiple protein types is required.”
If I thought he could understand it, I’d recommend Andreas Wagner’s Arrival of the Fittest to colewd.
Rumraket,
Here is an image that shows some of P53 binding pathways. I chose this protein because it is implicated in many cancers and there is a lot written about it and its mutations that cause cancer due to failure in the apoptosis process. You will notice its regulation strategy is the same as HIF and Beta catenin as it is down regulated through a destruction mechanism.
colewd,
What I’m trying to get you to notice, William, is that no matter what you prefer to believe. No matter what I’d prefer to believe, no matter what anybody prefers to believe, it’s a fact that there’s lots of variability, with no ill effects, among individuals within a single species. You have agreed to that, though you seem unaware of the implications. This variability must be possible regardless of the details at the molecular level. Otherwise, the variability would not be there.
Now, it’s painfully obvious, that the variants are somewhat invisible to natural selection, since so many individuals make it to reproduction, and since the variability continues being present. We don’t see an end to it.
Wouldn’t you agree that given what we see, scientists have no doubt but to think that there must be plenty of “functional space” available for exploration?
Maybe there’s a huge functional space. At least that’s what the evidence points to. There’s no need to “hit” a beneficial mutations, it’s enough for neutral and semi-neutral to be available for variability to continue, right? Also, remember I mentioned purifying selection (as did Rumraket. That kind of selection keeps the worst negative mutations at bay.
It is true that proteins vary in how much mutation happens to them, and it’s true that one factor is whether they form complexes with other proteins. What is not true is that there’s proteins that cannot be mutated without any ill effects. That’s where you’re going wrong. It seems like you imagine the little room to spare in proteins belonging to complexes for some impossibility for change. As if those sequences could not be any different. If that’s so, well, there’s evidence that the complexes could be very different. Here’s where we explore proteins across lots of species and we find that, regardless of the complex formed, there’s lots of variability even among the most conserved sequences, and they still do the very same job!
Well, no, that’s not a reason for the Darwinian mechanism losing steam as an explanation of life’s diversity. The real problem is not that the mechanism is wrong, but that life’s diversity is not just about the Darwinian mechanism. There’s also the very thing I’m trying you to get and understand: phenotypic variation that is not dependent on selection, but on, actually, being mostly invisible to selection. The problem is that the Darwinian mechanisms explains so much in such elegant terms, that scientists ignored the roles of other mechanisms in shaping life’s diversity for very long.
But we’re digressing.
… and?
Again, you claimed:
“There may be proteins where playing around is possible but they don’t live inside the cell nucleus where mutational sensitivity is high and binding to multiple protein types is required.”
It does not follow from:
“P53 … is implicated in many cancers and there is a lot written about it and its mutations that cause cancer due to failure in the apoptosis process … [and] its regulation strategy is the same as HIF and Beta catenin as it is down regulated through a destruction mechanism”
… that therefore:
“There may be proteins where playing around is possible but they don’t live inside the cell nucleus where mutational sensitivity is high and binding to multiple protein types is required.”
It.
Simply.
Doesn’t.
Follow.
Entropy,
I agree with your analysis here.
I think that functional space varies according to how much function and binding is required of the protein. It would appear to me that NCRNA’s would be very sequence dependent as they need to find specific locations to bind to. I mention nuclear proteins as they require multiple binding partners as part of their function. The image I posted for Rumraket shows some of the binding requirements for P53, a well researched oncogene. Since protein sequence is so large it would take an enormous functional space to allow a successful random search. Given the data that for 100 AA protein non functional space is 10^11 for just requiring folding and ATP binding (Szostak) a diversity built by random search seems very unlikely.
I agree with this it is just that the front end cannot be random.
Phenotypic variation is dependent on sexual reproduction that brings together two different sets of genetics. As we continue to 4 then 8 then 16 sets of genetics mixed together we get a natural increase in variation. In breeding we can control that variation by the animals we select.
Protein formation is just one problem with random searchs of the genome. Embryo development which requires very precise biochemical process is another.
The link in my previous post doesn’t seem to work so here it is again:
Hayashi, Y., Sakata, H., Makino, Y. et al. Can an arbitrary sequence evolve towards acquiring a biological function?
J Mol Evol (2003) 56: 162. https://doi.org/10.1007/s00239-002-2389-y
Rumraket,
I am not sure why you don’t see the connection. The diagram I posted showed 10 proteins that p53 needs to bind to. We also know that mutations to p53 causes cancer. You don’t see this as a problem for random exploration of the sequence space?
No, I don’t see how the problem with this particular protein can be extended to all of sequence evolution for nucleated cells.
That is the most, shall we say, ambitious extrapolation I have seen in quite a while.
Bill, what the fuck does this even mean?
What about if we do the random search part, then go through the results of that search and ensure that we keep what we’ve found which is an improvement and discard that which is not an improvement?
Does that increase or decrease the probability to reach a 100 AA functioning protein?
The actual problem is in how the data is acquired. The reason we know of “bad” mutations in P53 is because we noticed them as associated to cancer. It’s not that P53 could not mutate, it’s that we only know of the mutations causing ill effects because that’s the context of isolation.
Another problem you’re having seems to be your use of the word “random.” What do you mean here? Are you imagining evolution starting with completely random sequences each and every time? If so, why? If not, then what’s the problem? (Evolution doesn’t explore from random sequences each and every time. It builds from perviously “explored” sequence space. This should be obvious because there’s tons of available successful sequences to work on.)
OMagain,
The test we are talking about produces a protein that can bind to ATP. Performing a function is second step. Performing a function that can add a reproductive advantage is a 3rd step and may require additional proteins.
Entropy,
By random I mean the steps that change a gene prior to fixation. The gene that is changing could be functional or not.
colewd,
But did you understand this?
And this?
We don’t need Joe’s formula for that. Again, you’re approaching this the wrong way.
Well, in this case “sheer dumb luck” seems to be smarter than you in a very obvious way.
In most cases, palindromes with functions in DNA sequences don’t have to be perfect. They often have variation and are still functional. What’s often palindromic is not each instance of, say, a transcription factor binding site, but the consensus obtained from aligning several of those sequences and looking at the most frequent “letter” at each position.
There’s also the tiny little detail that actual “palindromes” are pretty short. It’s not impossible to write several palindromes with DNA, say, six letters long. It’s not impossible for those to occur, just by chance, in quite large numbers in a genome.
Check this out. Suppose that every base has equal probability in the human genome. The probability of any sequence six bases long, including any of the potential palindromes, would be:
p = (1/4)**6 = 0.000244140625
Now, that means that the sequence can be present in a six billion bases long genome (like the human genome):
0.000244140625 * 3.3e9 = 805,664.0625
805,664 times without taking into account the potential for variability!
With those kinds of numbers, the real problem is, from an analysis perspective, to distinguish the functional ones from the noise.
There was an article, some years ago, showing how easily binding sites in DNA can evolve.
Entropy,
I understand. The issue with mutational sensitivity is that it also is evidence to reduced functional sequence space. P53 is part of the group of proteins involved in programmed cell death (apoptosis). The mutation sensitivity is also evidence that this protein is important for apoptosis. This function is mission critical to multicellular organisms and does not look like it was formed from a step by step process.
Yet over the history of life may very unique proteins had to form including muscle, nerve, brain, eye, cell adhesion etc.
Entropy,
I understand. The issue with mutational sensitivity is that it also is evidence to reduced functional sequence space. P53 is part of the group of proteins involved in programmed cell death (apoptosis). The mutation sensitivity is also evidence that this protein is important for apoptosis. This function is mission critical to multicellular organisms and does not look like it was formed from a step by step process.
Yet over the history of life many very unique proteins had to form including muscle, nerve, brain, eye, cell adhesion etc.
Nope. Seems like you didn’t understand. Again, since the mutations were isolated from cancer cells, it only shows you a biased sample of mutations in P53, a sample biased towards deleterious ones. It cannot tell you if it has a reduced “functional sequence space”, because it doesn’t show you mutations that don’t have a deleterious effect in P53.
Again, the sample of mutations is biased.
It might be critical to some multicellular organisms. Not so much for other multicellular organisms. You have to remember that these proteins evolved through a large number of different stages within different organisms whose needs were neither as particular, nor as precise, as those of the organisms where they might be a tad more constrained (like ourselves). Of course the protein could not evolve step by step if it had evolved in humans alone. But once you think in actual evolutionary terms, you realize that the evolution of such proteins doesn’t present too much of a challenge if they evolved in organisms where less constrained activities were still advantageous. Think unicellular eukaryotes, multicellular-but-simpler eukaryotes (like sponges), etc.
Entropy,
I think we can infer a reduced sequence space from two facts. The number of proteins P53 must interface with and the fact it is sensitive to mutation.
You missed this point. If P53 was not mission critical to apoptosis then a mutation would not be so problematic. Knock out experiments of p53 in mice show tumors forming in the first 6 months of life.
Entropy,
The challenge is to get 24 proteins to interface together to form a complex function. How do you propose DNA organizes itself to do this? At 300 AA per protein thats around 24000 nucleotides to organize with 4^24000 possible arrangements. Evolutionary theory proposes random genetic change until some form of fixation in the population occurs. You may have 4^100 total trials in the history of living organisms to explore a subset 4^24000 arrangements of DNA. That says functional space specific to apoptosis is 4^23900 just to have a chance to create this one function randomly. All this to evolve one single function of eukaryotic cells.
So lets pull the cooption card and claim that these 24 independent proteins had performed another function and became modified through random mutations and all were able to create an apoptosis function. 24 sets of genes modified in such a way that they fit together and form a function. Lets say 10 AA of the co-opted proteins are modified. Thats 240 amino acid modifications that need to form the apoptosis function with 20^240 possible arrangements and our limited data tells us that some single substitutions cause failure.
I think there is room for skepticism here 🙂
Really? Not to the worshipers of sheer dumb luck… To them, it has creative powers beyond the possibilities of even human intelligence…
That’s why its worship demands exclusive devotion and blind faith…
Stop with the sheer dumb luck J-Mac. I’ve explained this too many times to you. It’s not mere randomness. Got it now or you prefer to continue looking like an uneducable imbecile? I think you’re not that stupid, but you seem determined to make me think otherwise.
It’s not sheer dumb luck that has powers beyond human intelligence, it’s nature that has powers beyond human intelligence. This should be obvious since we’re a product of nature ourselves.
Agreed with former, disagree with latter. Do you really not understand that if the sample is biased we cannot know if there’s mutations that don’t harm the protein? If a sample is gotten from cancer cells, we’re going to find harmful mutations. That doesn’t mean that no mutations can occur. Do you really not understand this?
I didn’t miss that point. Of course the protein is involved in apoptosis, and of course, since mutations that disable the protein promote cancer, it’s deletion will promote cancer. That doesn’t mean that the sample is any less biased. Here it goes again: the sample is biased!
Here you’re imagining that P53, exactly with the sequence is has in humans today, is the only one thing that could have done the job. You’re mistaking a historically frozen version for the one and only possible solution. If you look at some homologs to P53 you’d discover that there’s plenty of room for different versions of the protein that would do the job, just within the P53’s “sequence space.”
If you also realize that P53 is a DNA binding protein, you’d realize that there’s plenty of proteins to “choose from.” So, it was P53, but it could have been something entirely different. That makes your numbers even less impressive.
Not really. Again, your numbers are misinformed. We don’t need the protein sequence to be exactly what it is today. We don’t need the protein to evolve from random amino-acid strings in one go. Again, each protein is the result of a long history of successes. Evolution doesn’t start from scratch. The ancestors to P53 have been around for quite a while, meaning that it has evolved functions on top of other functions for quite a while.
So, a few things I hope you’re starting to understand:
1. The evolution of P53 didn’t happen in humans alone. It happened across a history of life forms where it was not as restricted as it is now in us.
2. That implies that the protein did not come from purely random exploration, but from the historical accumulation of success after success across generations of organisms.
3. Most of the mutations we know about are biased. That means that we don’t know about the protein’s “functional sequence space” from that piece of data alone.
4. Studying P53 homologs shows that the protein can be somewhat different and still do its work all right.
There’s always room for skepticism. That’s all right. Maybe the route towards these proteins is different to what we think. But, from the little knowledge I have, there’s no major challenge for natural processes to get where we are.
In any of your complains, there’s something missing in your understanding about evolutionary processes. You seem to think that P53 had to evolve within humans alone, forgetting that its evolution happened across organisms that weren’t humans. You seem to think that the precise sequence that P53 has is the one and only that it could have been, which, again, ignores that there’s versions of the protein that have differences with the human version and still do their job all right. You seem to think that P53, and only P53 could have done this job, but there’s plenty of proteins that bind DNA. So many other proteins could have been the source for P53’s particular role. You seem to think that each protein has to evolve from scratch from random assemblages of amino-acids, forgetting that evolution would naturally tend to work on what’s available, on the historically successful sequences available in nature.
So, no matter how skeptical you prefer to be, you should be able to understand that there’s nothing unreasonable about my position.
Entropy,
Were good here. You agree that the number of different functions a protein must perform the smaller the functional space is. I understand your point on bias and it is a good one but since we are not trying to make a precise calculation all I need is your agreement that the conditions that P53 operates under reduce its functional space relative to a protein that has a single function.
This is true however P53 has to also bind to other proteins along with DNA which is why the numbers are potentially so large. In order to make your point credible you need to limit the function of P53 to binding DNA. The problem is that P53 needs to bind with 10 different substrates and this limits the number of sequences that will perform this function.
I don’t think this is the case. I already made the point that P53 exists in mice. The problem is that because the sequence space is so large and proteins that need to bind with several other proteins have reduced functional space the, time and populations available to evolution are not adequate to perform these searches.
I don’t think this is the case. There are other sequences that would work but the problem is that given all that P53 has to do, it is exceeding unlikely that apoptosis could evolve through a series of random searches even if it was able to coop other proteins.
I mentioned cooption previously.
I think you are doing your best with a very difficult position. Cells operate on information that exists in a mathematical sequence. There are several layers to these sequences starting with DNA then alternative splicing then amino acid sequencing. Sequences are very good at making large variations on only a few set elements, they are terrible at finding functional sets through a random search.
Any calculation requires you to know the variables. The point of the bias is that you don’t really know the variables, and that you don’t know the “sequence space” that would work. You still made some obscure calculation that seemed to require that the search was completely random, rather than an evolutionarily-historic “search.”
Yet again, you’re assuming that these interactions have to evolve in one go from random trials. What about associations evolving across several organisms not requiring all of them to be present from the get go? Remember, P53 didn’t evolve in humans alone. It didn’t evolve in mammals alone either. The interactions of P53 with DNA and other proteins has evolved across a huge variety of organisms, and, again, there’s a plethora of successful sequences to evolve from. It’s not as if DNA and protein binding was that hard to evolve either.
Its main functions across organisms rely on DNA binding. I’m not restricting its function, I’m talking about its historical main jobs. Proteins that bind DNA abound. Proteins that interact with other proteins abound, proteins that bind to DNA and other proteins abound. My point was just that there’s many options. That knowing what’s present today doesn’t mean that such was the only potential source for such a function. My point was that the numbers are no longer impressive as soon as you realize we’re talking about evolutionary processes, where potential solutions abound.
Only if the exact complex was the only possible solution to the functions of P53 and its partners. You seem to extend your mistaken notions to every component of a currently existing process. But, again, options abound, and that makes those numbers much less impressive.
They might have a reduced “functional space,” that doesn’t mean that a single molecule would be the only one that could do the job. As protein interactions evolve, they might “freeze” into a single interface. That doesn’t mean that such interface is the only possible option. This, again, is a mistake coming from prior misunderstanding.
To understand this point, imagine that you studied a DNA-RNA interaction, and you found that the RNA sequence was “AUUCG,” and that the DNA sequence was the complementary sequence “CGAAT.” Well, that such is the present “interface” doesn’t mean that such is the only possible interface five-nucleotides long. It just means that’s the one evolved in the system. Well, same thing for P53 interactions. they’re mediated by some sequences, but that doesn’t mean they had to be precisely those sequences. Do you understand this?
Again, all I see is misunderstanding on your part, rather than any difficulty in my position. I know where to look for answers, and I can easily envision answers before even trying to search for anything. Given that, I doubt that my position has any problems.
False. Cells operate on biochemical / biophysical processes, which we conceptualize into mathematical sequences.
While you cannot know if cells would be terrible at performing a random search, you keep forgetting that it’s not a random search. It’s a search that has random elements, sure, but it builds upon historically successful prior “searches.” These are not random searches, these are evolutionary searches. That there’s an element of randomness to them, doesn’t mean that they can be conceived as purely random searches.
You mentioned co-option before. OK. Then why do you forget and then write, time and again, “random searches”?
Entropy:
Bill has been reminded of this dozens of times by now, if not hundreds.
I’m not sure whether he simply can’t get it, or whether he won’t allow himself to. Either way, it’s a huge handicap.
It’s belief.
Always a huge handicap, when not justified.
Glen Davidson
keiths,
Maybe if we put that by itself it will be a tad more noticeable:
These are not random searches, these are evolutionary searches.
Entropy,
Am I incorrect in assuming that evolutionary searches start with random change?
Do you understand the difference between random change and random assemblages of amino-acids?
Entropy,
No. I started assuming there is lots of functional space but the sequence space is larger enough to be problematic. We started with Szostak’s estimation for a single function of binding with ATP of 10^11 with 20^80 possible arrangements.
Not when the measured probability of binding to a single small molecule is 10^11.
If the functional space is reduced by having to bind to a second molecule you are moving to more improbable from an improbable number (10^11). We are not talking about 80 AA or single function. We are talking about hundreds of AA and over 11 functions.
I am not sure why this positioning is important.
We know of a long term experiment which resulted in citrate being consumed in an oxygen environment. We also know that the experiment started with a functional set of enzymes and ended with the same functional set.
Entropy,
They are purely random until they are fixed in the population. If I look at 5 beneficial genes and we assume evolution occurred then we could describe those 5 genes as the result of a random search even if that search started from a previously functional gene.
You’ve been cautioned about the word salad before, and you’ve been given tips on how to avoid it.
colewd:
Bill,
You know from experience that you are terrible at this stuff — terrible at anything involving science or logical thinking. Why, then, do you treat mangled thoughts like the above as if they had value?
Your first thought, when you find yourself disagreeing with people who are far more competent than you, should be: “What am I misunderstanding here?”
As I’ve said before, while it’s logically possible that you are right and that the competent people are all wrong, the probability of that is vanishingly small. Start with the assumption that you are wrong, and try to figure out where your mistakes are. In the unlikely event that you are actually right, it will become apparent.
As things stand, you assume that whatever pops into your head has value, and we waste a huge amount of time correcting these brain farts. Be realistic for a change.
Which is still wrong precisely for the reason I mentioned: because that’s the proportion in random assemblages, not from historically preselected sequences, with a huge range of them to continue evolving from.
Again! That might be true of random sequences. In life forms today, we have plenty of molecules that are already preselected that bind to several small molecules. It’s not a random search from scratch! Do you really not get it?
Again, no. We’re talking about already evolved, preselected, molecules that have succeeded through eons of molecular evolution. You’re not getting this. The molecules have been selected, recombined, spread, varied, etc. This is not a purely random search. This is not a search from scratch!
We are talking about a plethora of already successful sequences available for evolution to occur. Not about random molecules from scratch!
Nope, we’re talking about a plethora of already successful molecules, each existing in many slightly different versions, with somewhat different activities, that are available for evolution to occur. These are not random molecules. These are evolutionarily successful molecules. They have a huge history built-in.
No, we cannot call it a random search because those molecules have already succeeded in many ways, among them in solving many of the “problems” you seem to think they’d need to solve from scratch and in one go.
The way it works can only be called an evolutionary search. You seem to be using the term “random search” as an excuse to avoid understanding.
Bravo, Entropy! However make sure you keep a link to this comment handy, you may well need to copy and paste from it yet…..
When I ask colewd about how the world must be if his misunderstandings are reality he complains I’m creating a straw-man.
But we see evolution around us. At the very least we’re all familiar with the yearly flu vaccine. Yearly. And anti-biotic resistance is more of a problem every day.
If it’s not sheer dumb luck (ht J-Mac) then what is it? When and how does the designer operate? If proteins can’t ‘evolve’ where do they come from? Can we start with an empty box then somehow cause the “designer” to put some novel irreducibly complex biological artefact in it? Why not? We know it has to happen millions of times a day just to keep bacteria going! Is the designer aware of us so it will always act in a way to be invisible to our probing? How come it’s me asking these questions and not colewd or J-Mac? Perhaps I should go get me some of that sweeet Templeton money….
OMagain,
Thanks. I try.
And that would only be true if there were only one “right” sequence waiting to be found. Whereas there is evidence (Szostak etc) that protein sequence space is rich in functional proteins.
If you search for the article on google scholar, then click on citations, you get quite a bit of interesting articles. I then selected for “since 2017.” Wow. This field moves really fast. Check this one for example.
Alan Fox,
Do you consider binding ATP to be a protein function in itself?
Entropy,
This is based on the faulty assumption that binding to substrate A has anything to do with binding to substrate B. When mutating to substrate B the sequence is still facing the large void of non functional space.
Were not binding to small molecules were binding to large proteins. P53 has to evolve surfaces that bind to all 10 in order for apoptosis to work. Each protein is much larger then ATP. P53 has been around along time as a molecule that performs apoptosis.
Granting their pre existence for the sake of argument, p53 had to evolve a sequence that would fold into a protein structure that would bind to all 10 proteins in order to gain the advantage of cell death. In itself this is a miraculous event without guidance.
Random change from a functional sequence. The odds are large that it walks toward non function.
An interesting observation in this context is that p53 sometimes acquires inappropriate binding activity and thereby gains novel oncogenic activity. Should we also take these to be miraculous events that required the help of our benevolent Designer?