Doubt Concerning The Selfish Gene

[Master Wayne]

Doubt Concerning "The Selfish Gene"

In the first chapter of "The Selfish Gene", when talking about replicating molecules, Richard Dawkins writes:

"Can we reconcile the idea that copying errors are an essential prerequisite for evolution to occur, with the statement that natural selection favours high copying-fidelity? The answer is that although evolution may seem, in some vague sense, a 'good thing', especially since we are the product of it, nothing actually 'wants' to evolve.”

I think this is a very interesting question, but I don't get the answer. If high copying-fidelity is a characteristic that is selected, wouldn't molecules, with time, present less and less variability, putting an end to evolution?

 

[Fredrik]

It would only end if we get to a point where there are no copying errors at all, or all copying errors lower the offspring's chances to survive and procreate. The former is impossible because of quantum physics. There will always be some non-zero probability of an error. The latter is only possible if the environment has stopped changing as well, if there's no climate change, no change in food supply, no new threats from predators, bacteria, viruses, parasites, etc.

Edit: My answer is about the evolution of animals. I see that the question is about self-replicating molecules. The main ideas of my reply still apply, but the examples of changes in the environment are of course pretty bad if we're talking about molecules.

 

[Borek]

Actually some level of error is necessary for survival - as Fredrik already signaled, environment is changing. If you block the changes, perfect replicator will sooner or later hit the wall as it will be not able to adapt, and it will be eliminated.

 

[Monique]

See my signature

 

[atyy]

Actually some level of error is necessary for survival - as Fredrik already signaled, environment is changing. If you block the changes, perfect replicator will sooner or later hit the wall as it will be not able to adapt, and it will be eliminated.

How about things like recombination? Would you consider recombination to be a copying error? It would seem reasonable to do so, although it usually isn't. Nonetheless, it seems to provide another way to generate diversity even if DNA replication (in the more usual sense) were perfect.

Or how about adaptation through non-genetically inherited plasticity like VDJ recombination or learning and cultural transmission?

 

[Master Wayne]

I'm sorry if my reasoning is a bit too simple, but I'm new to studying evolution (although not new to evolution itself, I've been doing that for quite a while). You guys make some interesting points. But it just seems to me that if we have two molecules A and B, and if A makes identical copies of itself and B rarely does, then after a while the predominating molecule in the pool will be A: high-copying fidelity is selected. And now I understand what you guys said about variability never reaching zero, but wouldn't you agree that it would be at least smaller? I mean, wouldn't variability after ten years be quite smaller than it was when the pool molecules first started replicating? We'd have 80% of A molecules and 20% of B, C, D, etc.

 

[Monique]

if A makes identical copies of itself and B rarely does, then after a while the predominating molecule in the pool will be A: high-copying fidelity is selected.

Why would superior copying fidelity be selected? Genetic variety is beneficial for the gene pool. High fidelity polymerases (copying machinery) do exist, but that's in response to extreme environmental pressure: http://en.wikipedia.org/wiki/Pfu_DNA_polymerase

 

[atyy]

I'm sorry if my reasoning is a bit too simple, but I'm new to studying evolution (although not new to evolution itself, I've been doing that for quite a while). You guys make some interesting points. But it just seems to me that if we have two molecules A and B, and if A makes identical copies of itself and B rarely does, then after a while the predominating molecule in the pool will be A: high-copying fidelity is selected. And now I understand what you guys said about variability never reaching zero, but wouldn't you agree that it would be at least smaller? I mean, wouldn't variability after ten years be quite smaller than it was when the pool molecules first started replicating? We'd have 80% of A molecules and 20% of B, C, D, etc.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3628614/ discusses some of the issues, including the points that Fredrik and Borek brought up.

 

[Fredrik]

Why would superior copying fidelity be selected?

The new molecule is more likely to be able make copies of itself if it's an exact copy of the original (which is able to copy itself) than if it's a version of the original in which some random change has been made.

 

[atyy]

One thing in which there is very little diversity is the genetic code itself. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3293468/ discusses why this might be so.

 

[Monique]

The new molecule is more likely to be able make copies of itself if it's an exact copy of the original (which is able to copy itself) than if it's a version of the original in which some random change has been made.

There still would need to be a selective advantage, otherwise a Hardy-Weinberg equilibrium will be reached. With a slight mutation rate the polymerase is still able to make sufficient copies to maintain it's existence.

 

[ChrisJA]

In the first chapter of "The Selfish Gene", when talking about replicating molecules, Richard Dawkins writes:

"Can we reconcile the idea that copying errors are an essential prerequisite for evolution to occur, with the statement that natural selection favours high copying-fidelity? The answer is that although evolution may seem, in some vague sense, a 'good thing', especially since we are the product of it, nothing actually 'wants' to evolve.”

I think this is a very interesting question, but I don't get the answer. If high copying-fidelity is a characteristic that is selected, wouldn't molecules, with time, present less and less variability, putting an end to evolution?

I think the book is wrong. There is a scale. I think if you plot the rate of copying error vs. survival over time there is a peak in the curve. A perfect copying system with zero error would have good short term survival but very poor long term survival (because the environment might change.) On the other hand poor copying ability would not work so well either as most offspring would die.

There is likely an optimal error rate.

 

[mishrashubham]

I think the book is wrong. There is a scale. I think if you plot the rate of copying error vs. survival over time there is a peak in the curve. A perfect copying system with zero error would have good short term survival but very poor long term survival (because the environment might change.) On the other hand poor copying ability would not work so well either as most offspring would die.

There is likely an optimal error rate.

I see it somewhat that way as well (although I don't think the book is wrong, depends on the intepretation though). Natural selection favours that which survives to have another go at replication. A harmful mutation may lead to a nonfunctional/non-replicating offspring molecule, but seemingly harmless mutation might just be that thing the molecule requires to survive in a different environment sometime in the future.

 

[ChrisJA]

I see it somewhat that way as well (although I don't think the book is wrong, depends on the intepretation though). Natural selection favours that which survives to have another go at replication. A harmful mutation may lead to a nonfunctional/non-replicating offspring molecule, but seemingly harmless mutation might just be that thing the molecule requires to survive in a different environment sometime in the future.

The book said (I think) that natural selection favors exact replication. I said THAT was wrong, not the entire book. I think that exact replication disallows any kind of evolution.

 

[Ygggdrasil]

Exact, error-less replication may not be possible. There are many sources of mutations from replication errors by DNA polymerase to mutations induced by environmental conditions (e.g. UV radiation) and many of these are unavoidable.

Even if organisms could replicate without error, there would be the issue of diminishing returns. Obviously, a replication and repair system that allows too many errors would be detrimental as most progeny produced by such a system would have deleterious mutations. However, reducing the mutation rate has some trade offs, for example, in the amount of resources required for making various DNA repair enzymes and in the reduced speed of replication. At some point the marginal gain in fitness from a lower error rate would be counterbalanced by the great increase in resources spent on maintaining a low error rate.

Many have argued that it is this "cost of fidelity" that sets the optimal mutation rates of organisms at a non-zero value, not the evolutionary need to generate variation within a population (for example, see Sniegowski et al. 2002. The evolution of mutation rates: separating causes from consequences. Bioessays 22, 1057).