Yes I read them. And unlike you I understood them and therefore know they have the same flaws as all the other crap you people spam with. What’s amazing to me is you seem entirely unable to even conceive of what those criticisms might be.
Think I’m wrong? Give it a shot. Tell me first of all, taking the cell article as an example: What function did they discover?
Second, what are the usual criticisms of these types of studies from proponents of junk-DNA, and how did the authors take those into account? What did they in actual practice do to show how these transposable elements differ from junk?
Which squares well with @John_Harshman’ s claim at 187 that only 10% of the genome is functional. Don’t think it’s true of course, but this seems to be the experts consensus view here at PS.
I told you exactly what I did. I used an extremely lowball estimate of 1% for functional genome size.
The majority of mutations in the coding region will be amino-acid changing. Not all, but most. None of this is going to ultimately change the inevitable result of fitness decline. If you think you have a set of parameters you can defend which don’t show fitness decline, it’s long past time for you to present them.
The 7.2 figure was 10% of the 72 figure that UP/PP used. The genome length he used was 1 billlion base-pairs. Now he is using, in his later simulation, 10 million base-pairs, all non-junk. 1% of the 72 figure is 0.72.
In any case I think I was wrong to think that if 3 billion bases suffer 72 mutations per generation, we can use that figure for a genome of length 1 billlion. The appropriate way to respond to the limitation of Sim (to 1 billion base-pairs) would be to reduce the 72 to 24, and then take the resulting reduction in fitness as 1/3 of what would occur in a genome of 3 billion.
Allowing for 90% of the genome being junk then we should reduce the 24 to 2.4. In that case the genome length would be 100 million base-pairs. When reduced to 10 million, for faster simulation, one should use 0.24 mutations per generation, then take any reduction in fitness as projecting that we would have 10 times as much in a non-junk genome of size 100 million, or 30 times as much in a genome of 300 million non-junk sites.
So logically, you’re saying you think that 90% of mutations are strictly neutral and do nothing to fitness. This is exactly what Rumraket claims nobody thinks.
I don’t think that would work because selection dynamics will work differently with different mutation rates. The whole point of keeping the same mutation rate was to keep the mutation load dynamics accurate.
So you’re saying the exact thing that Rumraket confidently asserted nobody thinks.
This cannot possibly be a serious suggestion! You think we should take the accurate mutation rate for humans (72-100 per generation), and reduce it down to 0.24 (basically 0 mutations per individual since you cannot actually have less than one mutation). And this is supposed to be an accurate simulation? I think not. The only way to keep accurate simulation dynamics is to keep the same number of mutations per individual. It’s not about the genome size, it’s about the per-individual mutation rate, the DFE and the population size.
It actually occurs to me that all this nonsense is unnecessary if we’re assuming a functional genome of only 1%. In that case, the computational load problem is already solved without the need to further divide by 3 again. So I went ahead and retried the simulation with a genome size of 30 million bp (“full size”). Same result. Fitness decline to collapse. Junk DNA does not get you out of the GE problem. It actually makes it worse because if you are going to ignore most of the genome and only focus on the 1%, then you have to correspondingly use a DFE which is representative of that 1%, and that means a higher mean deleterious selection coefficient.
And yet he’s apparently also assuming that All 7.2 mutations are non-silent changes in protein-coding sequences. Serious misunderstanding.
I doubt it. Can you cite anything there? Considering that the bulk of mutations are transitions and that 3rd position transitions are all silent, I’m suspicious. I suspect too that the bulk of non-silent mutations are strongly deleterious and will not rise to an observable frequqncy.
LOL. So you made it up to get the conclusion you wanted. Hilarious. Try about 1.5% beneficial. You know, like this recent study that you hate so much, found:
Actually let’s be extremely generous to you and set it to 0.8% then.
Paul I think you should learn what a simplifying modeling assumption is. That’s really just what is going on here. It doesn’t mean we think everything we put into the model is strictly reality. When the model disagrees with reality, you should change it.
Joe Felsenstein is just giving you suggestions for how you can get simulation results that don’t disagree with reality, not giving you a lecture on how/why mutational effects actually distribute in junk-DNA.
The piece accurately reports that Cech doesn’t share the consensus view here at PS according to which most of the RNAs derived from the so-called « dark matter » of the genome is simply functionless transcriptional noise.
That does check out. From what I can find, estimates are that about 60-70% of DNA mutations in codons result in amino acid substitutions. Which does make sense when you look at codons.
Looking simply at the number of possibilities, only 8 amino acids have codons with fourfold degeneracy at the 3rd position making all mutations there silent, leaving 15 without.
Of the 15 without, 12 have twofold degeneracy so now half of possible mutations at the 3rd position aren’t silent. 2 have no degeneracy (all mutations are non-silent), and 1 has threefold.
That still makes the majority of codons not have the 3rd position mutations result in mutations.
Further still, at the first and second position all mutations result in amino acid substitutions.
The only thing keeping it down is really that the code is structured so the most likely mutations (transitions) are typically also the bases found in codons with less than fourfold degeneracy. To minimize the expected deleterious effect of mutation, when mutations do cause amino acid substitutions, they are biased towards more physico-chemically similar amino acids.
This does raise a question about why the code is structured in this way on creationism. On evolution it makes sense that selection has shaped the code to minimize the effects of deleterious changes and maximize the probability of adaptation.
But on YEC God presumably didn’t create life to evolve by random mutations occurring in blindness to their future effects, and creationists typically assert life was created with frontloaded diversity that somehow simply segregated out in different lineages.
Then why is there a genetic code with this structure in it? Did mutations happen before the fall, but God intended them to happen? Were they random back then too? What is the creationist hypothesis here. God created life intending it to stay in the garden, not to evolve. Then the fall happened, and… a magical curse inside an apple structured the genetic code to reduce the effect of mutation? God created the code to do that, beforehand, just in case? But for no reason whatsoever because life is going extinct anyway, the code can’t prevent that?
So Felsenstein is trying (but not succeeding) to stack the deck in favor of evolution by deliberately making the parameters unrealistic.
You stated nobody thinks that mutations in junk are strictly neutral, but Felsenstein apparently disagrees. You two can hash it out, but I’m not going to be gaslit into thinking you’re saying the same thing. Either way though, fitness still declines, so I don’t have a stake in the game.
