Do you think this data has a chance to show us the functional information of MYH7 is less then alpha actin?
These data show that you are grossly understating the lack of sequence conservation of MYH7.
Myosin has far more functional information than actin. These two proteins demonstrate the silliness of trying to reduce it to a formula that doesn’t correspond to anything resembling functional information.
Why didn’t you answer my other questions, Bill?
What happens if we split a conventional myosin II, such as the one from MYH7, in half? Do those halves have equal functional information, or are there orders of magnitude difference between them?
1 Like
I am certainly briefly read the paper and may have some misunderstanding. Can you summarize why you think Lynch gets different results then Behe? Are you claiming it has nothing to do with how they treat gene duplications?
Clearly given Szostak and Hazens method for estimating functional information your claim that MHY7 is functionally more complex then alpha actin would be is confirmed. Whats left to discuss?
Why you insist that @gpuccio’s method, which requires the absurd assumptions that sequence conservation and length are meaningful proxies for functional information, is valid, since it gives the opposite result.
1 Like
Again, if we can’t define the difference between designability and evolvability, then it doesn’t even make sense to talk not evolvable ot not designable, because we can’t even tell the difference.
2 Likes
Hi John
If we assume Hazen and Szostak method is sound it validates your intuition.
How many substitutions have we observed in MYH7? Less than 500? How many are actually deleterious 10% or more? If we measure alpha actin the same way (human variation only) we get around 10 substitutions.
Regardless there is no chance this number would be big enough to over through your intuition.
I disagree with your assertion that length and conservation are not valid indicators of functional information and the comparison of MYH7 and Alpha Actin (using Szostak and Hazen formula) did not confirm your assertion.
Since we are observing the past can we change the definition to inferring design or the work of a mind vs inferring evolution or the work of natural processes and chance.
Bill why are you not able to understand that conservation is not a contradiction of evolution, but a predicted outcome? We expect there to be some protein sequences that become strongly conserved over time. That there will some times be proteins that gradually evolve to that state where basically any change has become deleterious.
The fact that some proteins become strongly resistant to change does not somehow mean there are no other proteins that can change.
Nobody has said that ALL proteins will be able to just drift around endlessly in sequence space, and there is no requirement that this should happen if evolution is true.
Your whole argument here is a non-sequitur. α-actin being 100% conserved in some vertebrate clade doesn’t mean all proteins are like α-actin(obviously), nor does it mean α-actin could not evolve to that state.
The article I linked earlier about protein sequence divergence explains in great detail what factors influence how constrained (in degree) protein sequences become through evolution. They explain that, for example, one of the strongest factors that affect the rate of protein evolution is the abundance of the protein in the cells in which it is expressed. At increasingly high levels of abundance every residue in the protein is increasingly exposed in frequency and intensity to some biochemical interaction, causing even very tiny biochemical effects to manifest more strongly. They are even able to derive a mathematical relationship between abundance and selection.
An evolutionary explanation for the gradual evolution and eventual conservation of a protein like α-actin is actually very simple. To explain how α-actin became so conserved it doesn’t really matter how it’s ultimate ancestor originated, whether that is de novo from non-coding DNA, by fusion and recomination of fragments of other proteins, or by even older duplications and divergence. Heck, it could even in principle have been designed and created, it doesn’t matter. It doesn’t matter because what we are trying to explain now is how a protein that is 100% conserved could have gradually evolved from some ancestral protein that was not, at least initially, as strongly constrained in sequence.
Do you follow?
A very rough outline of such an explanation would go something like this:
So, through some means, the protein’s ancestor originates and performs one or more different functions. This actin-like ancestor undergoes duplications, and these then diverge, then some of the diverging proteins acquire new functions(one of which is the function α-actin has.) This α-actin becomes extremely highly expressed because this is beneficial, and as a consequence of it’s abundance and it’s function becomes entrenched so any further change becomes strongly deleterious.
Meanwhile the pre-neofunctionalized duplicates that didn’t gain the function of α-actin, continue to diverge (because they are not as abundant and don’t perform the same function, therefore are less constrained), and still others acquire functions that are a bit more constrained while not being completely constrained(other actins).
There is a spectrum that goes from almost no constraint, to 100 constraint, and it depends on the function and the circumstance.
The problem with Behe’s argument is he’s basically pretending that all proteins are like α-actin has become(it did not begin this way, it became this way). But it’s just plainly and obviously false that all proteins start out being like α-actin, and a protein having become 100% conserved through evolution doesn’t mean that is how the duplication and divergence of all other proteins must begin.
Obviously
Obviously
Obviously
4 Likes
Sure. The short answer is that Behe deliberately created a model that was designed not to work, whereas Lynch attempted to create one that reasonably accurately reflected reality. But to the specifics:
Lynch’s model assumes intermediate mutations will be neutral. Behe assumes they will produce a non-functional protein and, therefore, once any mutation other than those required for the new function occurs that particular gene will be removed from the pool of genes that could develop a new function, since even if one of the target mutations subsequently occurs the protein will remain non-functional.
While Behe assumes that duplicate copies of the gene are already widespread thru the population, Lynch starts from a more realistic and relevant starting point, in which only a single copy of the gene exists. Rather than facilitating his model, however, as Lynch writes: “With many fewer initial targets for mutation, and the vast majority of new duplicates being rapidly lost by genetic drift, this starting condition imposes a much greater challenge for the evolution of a new gene function than that assumed by Behe and Snoke.” So when you claim that Lynch’s finding that neofunctionalization is far more likely than Behe calculates depends on gene duplications being less frequent, you are simply wrong. This is actually a condition he imposed that works against his model, but which he felt compelled to include because it more closely reflects reality.
A further difference is that Behe assumed that only two specific mutations could possibly lead to a new function involving two mutations, whereas in Lynch’s model there are multiple possible two mutation combinations that could lead to a new function.
There may be other differences, but those are the main ones to my understanding of the papers.
7 Likes
Hi Rum
I think you logic holds together well. Where we differ is in how proteins like alpha actin and MYH7 originated.
-Was it designed in the current state we are observing?
-Could it evolve by random change and selection into the state it is in?
How would your argument change under the first condition?
As far a Behe and Lynch I think both models are not universal and depend on the protein type. I may be wrong but based on what you wrote I think you agree with this. To eliminate a possible edge to evolution a universal model is required.
Hi Faizal
He assumes they will be deleterious in most cases but it does not make sense that a single AA substitution will make the gene non functional. Where does he make this claim?
You are correct, that is my mistake. Neutral mutations are ignored in his model, but he uses a variable p which is his estimate of the number of null mutations (mutations that lead to a non-functional protein) to the number of mutations that could create a new function.
2 Likes
I don’t assume that it is sound and it’s not my mere intuition. It’s from my experience studying function, which you don’t have. YOU are the dilettante depending on intuition, which routinely fails when understanding biology. Please don’t engage in such silly projection.
I’m pretty sure you’ve observed none.
Why would 500 be a magic number?
So now you’re going back to denying the existence of penetrance? Why?
“How many are actually present in healthy people?” is the relevant question.
It’s hilarious that you are trying to discount real evidence that you won’t examine as mere assertion.
You haven’t looked for all of the relevant data yet, so you can’t possibly have used the formula.
1 Like
Thanks I appreciate that.
I don’t see how it would change anything about what I’ve stated above.
The question I was trying to address is whether a protein being strongly conserved somehow indicates it couldn’t or didn’t evolve. I just don’t think the protein being conserved is at all an indication of that for the reasons stated.
Just as a designer can design a protein to be under very strong constraints, a designer can also design one that has a very high tolerance to change. And likewise there are circumstances under which proteins can evolve to become entrenched and resist further change, and there are circumstances under which they can tolerate a lot and diverge continuously over extremely long timescales.
Given that both scenarios are in principle totally compatible with different proteins exhibiting the entire range on the spectrum of conservation, a protein exhibiting some particular level of constraint - no matter how strong or weak - can’t be a factor that can be used to discriminate between them.
That means we have to look at other factors to understand how some particular protein came to exist.
Yes, definitely. The strengths of the constraints operating on different protein sequences really do differ from one protein to another, and can even change over time as the environment in which the organism lives, changes over time too.
It would be a mistake to try to argue against the principle that there must be some sort of edge to evolution. I think we can know from reason that there must be such an edge.
Again the problem with appeals to the edge is that no concrete example is known of something that has been shown to actually lie beyond this edge. It’s not that there could not be such a thing, at least in principle.
4 Likes
No. You cannot arbitrarily change definitions and claim any standing as science. You MIGHT argue that a definition should be refined or modified, if there is a convincing case that such a change is needed.
2 Likes
Are there definitions in science for evolvability and design ability.
I think there is a very strong case that structures like the bacterial flagellum and complex proteins are beyond the edge of evolution as even with Lynch’s assumption these structures are very unlikely to form in a population.
Wrong.
2 Likes
It seems to me that if there really were, you would have done it by now.
2 Likes