Mutations are for living beings like the rust for a car. When you buy a new car, it will start to accumulate point of rust. At the beginning, rust will be invisible and will have no effect. But over time, the rust points will accumulate to a point where it will become visible, without compromising the proper functioning of the car. But as the rust continues to build up inexorably, there will inevitably come a time when the car will fall apart.
Lynch has shown that GE is inescapable for small population.
https://www.jstor.org/stable/2462976
Lynch again has shown that GE is a real threat for public health
https://www.pnas.org/content/pnas/107/3/961.full.pdf
Easy. Since in this type of analyses, fitness corresponds to reproductive rate, the increase in fitness seen in your curve is most likely explained by loss of function. IOW, by devolution or GE. This is what Lenski’s LTEE showed.
1 Like
No, it’s actually completely obvious that considering the total space of all possible beneficial and deleterious mutations, the ratio must be 50/50 beneficial to deleterious. Because for any imaginable mutation that is deleterious, it’s reversal is beneficial. And any mutation possible has a complementary reversal. Any duplication has a corresponding deletion, any substitution A->G can be inverted G->A.
This is logically unavoidable.
Now it may be the case that for many phenotypes they have climbed to somewhere close to some local adaptive peak, and so from their perspective a majority of new mutations available to them are reproductively downhill again, but should they occur, their reversal will become available again.
1 Like
Yeah that’s a nice analogy, except natural selection doesn’t exist in that analogy, and seems to have been declared powerless to oppose the rusting process by fiat. In exactly the point of contention (whether NS can oppose the rusting process by removing deleterious mutations and fixing beneficial ones), the analogy simply begs the question.
Nevertheless, observation of real populations of living organisms, as opposed to rusting cars, show continued fitness increases, which means the real world is not like the rust-analogy: Natural selection is in fact capable of opposing the buildup of deleterious mutations, by some combination of purging them from the population and by fixing actually beneficial ones that really do occur, quickly.
No, he’s shown extinction due to mutational meltdown becomes increasingly likely the smaller the population, not that it is “inescapable”. And by small populations he’s generally speaking of species with effective population sizes well below 1000.
Also, the straightforward implication is that the bigger the population, the less likely extinction due to mutational meltdown becomes.
Thus, because the effective sizes of natural populations are often on the order of a tenth to a third of the actual number of breeding adults (Heywood 1986; Lande and Barrowclough 1987; Briscoe et al. 1992), avoidance of extinction by a mutational meltdown over time periods of 100 or more generations appears to require the stable persistence of well over 1,000 breeding adults per generation.
And finally, “Genetic Entropy” =/= loss of mean fitness due to accumulation slightly deleterious mutations. Genetic Entropy is some nebulous, ill-defined term that has something to do with “damage” and “deterioration”, but isn’t actually the same as fitness.
No article you link with extinction due to loss of fitness by accumulating deleterious mutations is evidence of Genetic Entropy. As Price is kind enough to concede:
We’ve been through this. Fitness is contextual, and is not defined in the same way in every experiment. It’s not always a proper metric for what Sanford’s GE is talking about, which is ultimately not ‘fitness’ but information . Most of the time, degrading the information in the genome does result in a loss of fitness, but not always, and not if you create an attenuated branch that is narrowly well adjusted to a particular experimental environment (but which has lost information in the process).
2 Likes
So in fact GE has nothing to do with extinction due to loss of fitness by “inescapably” accumulating reproductively deleterious mutations.
Thus, no article that shows extinction due to loss of fitness can by that alone constitute evidence for GE.
Glad we got that settled.
2 Likes
Mutations visible by NS, whether deleterious or beneficial, are the tip of the iceberg. But beneath the surface is the continent of mutations invisible to NS. And it is only these mutations that are concerned by the rusting car analogy. I should have made this point clearer.
1 Like
But you did not invoke the analogy merely to explain what is understood by the concept of GE, you brought it up in answer to T_aquaticus´ asking why Sanford believes GE will inevitably lead to extinction? The implication being that Sanford believes this is so because natural selection is powerless to stave off or counter the rusting process. More specifically the question being asked is why Sanford believes this is unavoidable.
And that is the point of contention. It seems to me, sticking with the analogy here, that when most of the car has rusted, the opportunity for reversal must be much higher than if most of the car has yet to rust. For a car with very little rust on it, it will be unlikely for a new mutation to hit an already rusted piece and restore it back to mint condition. But as the rust builds up, opportunities for reversal increase, which can be fixed in the population by natural selection. Not to mentions that it is also possible for mutations to strengthen other parts of the car, to compensate for those being weakened by rust.
Ah, OK I’ve seen somebody try to argue this before. But you are talking about back mutations. And your argument does not work because back mutations are rare. They are certainly not equally common, and simple probability will explain to you why that’s the case. You are forgetting that the first mutation (A ->G) was random, but the second one (G ->A in the same spot) now has a low probability. You are talking about a mutation randomly happening to hit in that same exact spot (lightning striking in the same place twice), and not only that, but now you’ve only got a 25% chance of reverting it back to the original nucleotide (A, T, C or G). EDIT: Oops, thanks Mercer! Actually certain nucleotide switches are more common, so it’s not as simple as 25% (but this does not change the point being made here). So naturally back mutations are not common and cannot make up for all the deleterious ones that are happening.
In fact, random back mutations are so improbable that it would make more sense to assume they are not random at all, if you do find that they are commonly taking place in a particular situation. It would look more like a non-random repair mechanism.
I also think that @Giltil and @PDPrice should try to agree among themselves about whether they think GE implies life could not have existed for billions of years.
@Giltil appears to think that GE merely implies that the diversity of life could not evolve: Design and Nested Hierarchies - #395 by Giltil
Giltil: No, Sanford mathematical analysis doesn’t implies life could not have existed for billions of years. Rather, it implies that life could not have evolved through the RV/NS mechanism. And so we have to look for another explanation.
… while @PDPrice has given links which argue that GE really does show life could not have existed for billions of years. For example this: Genetic Entropy and Simple Organisms
Dr Sanford argues that humans could not possibly have been around for tens of thousands of years (let alone millions, or billions if one considers our supposed evolutionary animal ancestors) because, at the current rate of mutation and the number of generations that would have occurred, we should have already become extinct.
That is an interesting discussion, but I would say it is secondary to the topic of this thread, which is “What are the fundamentals?”
So far I’ve been surprised to find so little discussion and/or understanding of those basic fundamentals (not intended as an insult to anyone).
I’m not making an argument about the frequency of occurrence of the fitness effects of mutations. At least, I’m not claiming that the frequency of occurrence(or opportunity) will ever rise from being biased towards deleterious, to be equal.
That’s why I began by saying “considering the total space of all possible beneficial and deleterious mutations”. To elaborate a bit, we can imagine the space of all possible viable genotypes for all possible living organisms with DNA, and then for each of these genotypes, we can imagine the total space of all possible mutations that could occur to them. And then for each of these mutations, for all of these genotypes, if the mutation is deleterious, we put it in the list of deleterious mutations, and beneficial ones in the list of beneficial mutations. What would be the relative sizes of each of those lists? They’d have to be exactly equally big.
I’m saying that their total numbers logically have to be exactly equal. There may be many good reasons for why the frequency of occurrence is biased significantly towards deleterious mutations, but that’s not a point I’m disputing or arguing against.
The real point of contention is if the deleterious bias is enough to render beneficial mutations too rare to counter the effect of deleterious ones. And it appears, from real experiements, that they are not so rare at all, since in fact many experiments have demonstrated continuous fitness increases.
And your response seems to have been to start talking about something other than fitness. That in fact when on the topic of GE you’re talking about some sort of measure of information (or function?), not fitness.
In which case any evidence you have for the preponderance of deleterious over beneficial mutations isn’t evidence of GE, nor the inevitability of extinction due to their accumulation.
1 Like
But that was the whole point of the discussion where you introduced this 50/50 concept (the frequency!). The back mutations you’re talking about here are only ‘beneficial’ in the context that they are reversing the effects of a deleterious mutation that has already occurred. That’s not forward progress. They in no way are even a factor worth considering in any fitness distribution. And thus the point remains–the vast majority of all mutations are deleterious (even the so-called neutral ones).
Which appears to remain an irrelevant point to the question of Genetic Entropy, because it’s apparently not about fitness at all. You’re just using the fitness effects of mutations as some sort of proxy for it, up until the point where it can be shown some population is exhibiting fitness increase, in which case you’ll change the subject and start talking about “deterioration” of some sort of information-ish property you haven’t defined or quantified. And insist that’s the real Genetic Entropy that someone in some future instance will “delve into a bit”.
2 Likes
Yes I’ve been forthright about the fact that it’s not easy, or maybe even possible, to quantify immaterial information content directly. You appear to be forthright here in accepting that my statement about mutations was accurate.
Which appears to remain an irrelevant point to the question of Genetic Entropy,
No, this fitness distribution is very much fundamental, and that’s why I included it here. So far I’ve seen no way out of this problem presented by anybody. Your appeals to attenuation, showing a localized (and ultimately temporary) increase in fitness do not overturn this problem. They are still reductive evolution. You are correct to point out this issue of terminology, but it’s not an issue that I, or even Dr. Sanford, are responsible for creating. It’s a by-product of the way fitness is measured and defined in general. In most cases there’s no difference because fitness lines up with function. Sometimes, though, lower function can be seen as higher fitness. It’s the function we’re concerned about, because that’s what keeps us all alive.
Your counterexamples might be compared to a situation where we are defining ‘fitness’ for cars as “the car doesn’t get stolen”, and then you point to a rusted-out car that’s on the brink of total breakdown and remark “See, nobody stole it! It’s the fittest.”
According to “genetic entropy”, increase in fitness cannot be seen. Yet here it is.
But thanks for the handwaving I was expecting.
1 Like
Not at all the same thing. Lynch shows is a decline in fitness when selection against deleterious alleles is prevented by technology. (Incidentally, how is fitness measured here? Fitness in what environment? Apparently, in the environment before the technological innovations were introduced?)
I do not agree that GE is a coherent explanation for anything, but even I wouldn’t go that far.
1 Like