Coat color evolution--is there evidence for new mutations?

[T_aquaticus]

@Mercer

Figure 2 already shows that that coat colors are not discrete, indicating plenty of polymorphism as I would expect. That was why I asked you whether you were claiming that there were no dark regions in the desert.

The whole point of the discussion is about dark regions in the desert, so I’m not sure what you are going on about.

@Mercer

Are you aware of the number of loci that affect coat color in mice and other mammals that have active switching between black and yellow pigment to give the base agouti color?

I can’t list them off from memory, but I am well aware that there are multiple genes that can affect the timing and expression levels of eumelanin in mouse hair. This fact is also a big part of the argument I am making.

As a reminder, mutations in just one gene (mc1r) perfectly correlated with the black fur phenotype in the Pinacate population. Mutations in mc1r are also know to be one of the genes that affects fur color in lab mice, as well as in other mammal species (including humans). Those mutations are not found in the Armendaris population. It is assumed that differences in genes other than mc1r are responsible for black fur in the Armendaris population.

So why are these convergent phenotypes segregated by geography? Part of the answer could be a lack of gene flow between geographic regions. It could also be due to strong selection eliminating that allele across the large geographic distance between the two populations.

[T_aquaticus]

@Mercer

How likely would it be to eliminate alleles in subpopulations given these data?

For populations in close proximity? It probably wouldn’t happen. What about populations that are not in close proximity, such as those separated by 100’s of km?

Although there is substantial genetic structure between geographic regions, high levels of gene flow (or recent ancestry) connect populations within a region. Despite these high levels of local gene flow, color variation is strongly correlated with habitat color in most populations, suggesting that natural selection for substrate matching is strong in this species.
Local adaptation in the rock pocket mouse (Chaetodipus intermedius): natural selection and phylogenetic history of populations | Heredity

[Mercer (John Mercer)]

@T_aquaticus

The whole point of the discussion is about dark regions in the desert, so I’m not sure what you are going on about.

The whole point of this discussion is to see if you have evidence to support your assumption that the raw material for this rapid evolution was new mutation or if it is just as likely to have been existing polymorphism.

@T_aquaticus

I can’t list them off from memory, but I am well aware that there are multiple genes that can affect the timing and expression levels of eumelanin in mouse hair. This fact is also a big part of the argument I am making.

I don’t see how that helps the assumption you’re trying to defend.

@T_aquaticus

As a reminder, mutations in just one gene (mc1r) perfectly correlated with the black fur phenotype in the Pinacate population.

No, as a reminder, an allele of just one gene perfectly correlated.

You are asserting that it was a new mutation. I am pointing out that you have no evidentiary basis for doing so.

@T_aquaticus

Mutations in mc1r are also know to be one of the genes that affects fur color in lab mice, as well as in other mammal species (including humans).

You’re assuming your antecedent by using “mutation.”

Wild populations of both humans and mice are polymorphic at this locus and many others. That’s consistent with the top graph here:

And falsifies your hasty generalization from the tiny sample in the other paper:

@T_aquaticus

On top of that, the coloration was discrete:

In these populations, phenotypic variation in color was largely discrete

Note that in your zeal, you omitted the qualifier “largely.” The top graph shows that for a larger sample, variation isn’t even “largely discrete,” much less “discrete.” That’s the sort of variation we would expect to see with polymorphism at multiple coat-color loci.

[T_aquaticus]

@Mercer

I don’t see how that helps the assumption you’re trying to defend.

Let’s say that there were no black alleles in the pocket mouse population that spread out into the deserts of Arizona and New Mexico. Once they had established themselves, the populations at each of the two sites acquired new mutations that produced the black phenotype. Given the numerous genes in which mutations can produce this phenotype, it is very possible that new and independent mutations in different populations would occur in different genes.

Let’s say that the black allele was present in the original population that spread out into the desert. Chances are, we would have a mixture of different genotypes for the black phenotype at both sites. Therefore, you would expect to see different genotypes for black fur at the same sites.

What do we see? We see that the black phenotype in different and distant geographic have their own unique genotype. To me, that points to mutations that occurred at each site independent of each other.

@Mercer

You’re assuming your antecedent by using “mutation.”

Are you saying that the differences between alleles are not due to mutations?

@Mercer

Wild populations of both humans and mice are polymorphic at this locus and many others.

That wasn’t the case for the mc1r black allele at the Pinacate site.

[Mercer (John Mercer)]

@T_aquaticus

Let’s say that there were no black alleles in the pocket mouse population that spread out into the deserts of Arizona and New Mexico.

That’s the assumption you are making. It is not supported by any data.

@T_aquaticus

Once they had established themselves, the populations at each of the two sites acquired new mutations that produced the black phenotype.

Why is that more likely than existing polymorphism?

@T_aquaticus

Given the numerous genes in which mutations can produce this phenotype, it is very possible that new and independent mutations in different populations would occur in different genes.

No, it’s much less likely than existing polymorphism and founder effects, especially given that top bar graph in the figure above. I note that you haven’t commented on it. Why?

@T_aquaticus

Let’s say that the black allele was present in the original population that spread out into the desert.

OK, but which black allele for which locus??

@T_aquaticus

Chances are, we would have a mixture of different genotypes for the black phenotype at both sites.

No, because all the data in the first paper say that the black phenotype in Armendaris is a different locus.

@T_aquaticus

What do we see? We see that the black phenotype in different and distant geographic have their own unique genotype.

So? How does that exclude existing polymorphisms?

@T_aquaticus

To me, that points to mutations that occurred at each site independent of each other.

You haven’t explained why they weren’t alleles already present in the population.

Here’s something that I have been waiting for you to acknowledge: if this was a new mutation, shouldn’t it only differ from WT by a single residue? Why are there four substitutions, all of which change charge? What’s more likely: four simultaneous new mutations or a long history of selection?

@T_aquaticus

Are you saying that the differences between alleles are not due to mutations?

We’re discussing your unsupportable assumption that it was a new one. Who’s to say that the black allele wasn’t the WT in the founding population before it colonized the desert?

@Mercer

Wild populations of both humans and mice are polymorphic at this locus and many others.

That wasn’t the case for the mc1r black allele at the Pinacate site.

Absolutely false; look at Table 1 and Figure 3. That’s the third time you’ve been wrong about the actual data.

What about gene flow?

@T_aquaticus

A good scientist spends as much time trying to prove themselves wrong as they do trying to support their hypothesis.

Something you sure aren’t doing here.

[T_aquaticus]

@Mercer

That’s the assumption you are making. It is not supported by any data.

That’s the scenario I am presenting. I am not assuming, I am hypothesizing. There’s a big difference.

@Mercer

No, because all the data in the first paper say that the black phenotype in Armendaris is a different locus.

I don’t think you are understanding what I am trying to communicate.

I am setting up two scenarios and determining which observations we would expect to see in each. It is hypothesis A and B. Let me label them as such:

Hypothesis A: Let’s say that there were no black alleles in the pocket mouse population that spread out into the deserts of Arizona and New Mexico. Once they had established themselves, the populations at each of the two sites acquired new mutations that produced the black phenotype. Given the numerous genes in which mutations can produce this phenotype, it is very possible that new and independent mutations in different populations would occur in different genes.

Hypothesis B: Let’s say that the black allele was present in the original population that spread out into the desert. Chances are, we would have a mixture of different genotypes for the black phenotype at both sites. Therefore, you would expect to see different genotypes for black fur at the same sites.

@Mercer

Here’s something that I have been waiting for you to acknowledge: if this was a new mutation, shouldn’t it only differ from WT by a single residue?

If it is new then it should have less divergence than the brown genotypes, and it does.

Finally, the pattern of nucleotide variation observed among Mc1r alleles from the Pinacate site suggests the recent action of positive selection. Thirteen polymorphic sites are variable among the light haplotypes, whereas only one site is variable among the dark haplotypes (Table 1). The ratio of variant to invariant sites is significantly different between dark and light alleles (1/953 and 13/941, Fisher’s exact test, P < 0.01). The average level of nucleotide diversity among light alleles (π = 0.21%) is >10 times greater than nucleotide diversity among dark alleles (π = 0.01%). The reduced variability seen among the dark Mc1r alleles is the expected pattern if selection has recently fixed an adaptive substitution (26–28).
https://www.pnas.org/content/100/9/5268

Why are there four substitutions, all of which change charge? What’s more likely: four simultaneous new mutations or a long history of selection?

From the paper:

Several observations suggest that one or more of these four amino acid mutations (sites 18, 109, 160, and 233) are responsible for the light/dark phenotypic differences seen in the Pinacate population.
https://www.pnas.org/content/100/9/5268

It could be one adaptive mutation with the other three hitching a ride.

@Mercer

We’re discussing your unsupportable assumption that it was a new one.

You didn’t say new one.

You’re assuming your antecedent by using “mutation.”

If pedantry like the above is bothersome, then I would suggest we both stop doing it.

@Mercer

Who’s to say that the black allele wasn’t the WT in the founding population before it colonized the desert?

If that were the case then I would expect to see the black mc1r allele at both black lave sites. We don’t. We see different genotypes for the same phenotype.

@Mercer

Absolutely false; look at Table 1 and Figure 3. That’s the third time you’ve been wrong about the actual data.

Just in case you missed it from above:

Finally, the pattern of nucleotide variation observed among Mc1r alleles from the Pinacate site suggests the recent action of positive selection. Thirteen polymorphic sites are variable among the light haplotypes, whereas only one site is variable among the dark haplotypes (Table 1). The ratio of variant to invariant sites is significantly different between dark and light alleles (1/953 and 13/941, Fisher’s exact test, P < 0.01). The average level of nucleotide diversity among light alleles (π = 0.21%) is >10 times greater than nucleotide diversity among dark alleles (π = 0.01%). The reduced variability seen among the dark Mc1r alleles is the expected pattern if selection has recently fixed an adaptive substitution (26–28).
https://www.pnas.org/content/100/9/5268

[Mercer (John Mercer)]

@T_aquaticus

I am setting up two scenarios and determining which observations we would expect to see in each. It is hypothesis A and B. Let me label them as such:

Hypothesis A : Let’s say that there were no black alleles in the pocket mouse population that spread out into the deserts of Arizona and New Mexico. Once they had established themselves, the populations at each of the two sites acquired new mutations that produced the black phenotype.

If the mutations occurred after the lava flows, then the time between now and the time of divergence of all of the alleles, both black and beige, shown in the paper must be less than a half-million years. Much less, since the sample is so small and there are almost certainly other polymorphisms that are being missed.

Correct?

Hypothesis B : Let’s say that the black allele was present in the original population that spread out into the desert. Chances are, we would have a mixture of different genotypes for the black phenotype at both sites. Therefore, you would expect to see different genotypes for black fur at the same sites.
[/quote]

That’s not very realistic, given what we know from the house mouse and the gradient of color observed in the paper you just found.

Hypothesis C: The founding population was as polymorphic as house mice, with multiple alleles at multiple loci. This allowed them to inhabit multiple niches in the desert, even darker ones like shady, north-facing cliffs. Different black alleles and different black/yellow loci (such as Agouti, the first place the PNAS paper’s authors looked) are likely to have been selected on different flows because of founder effects.

This predicts that calculated divergence times for pairs of alleles will be longer than the age of the lava flows.

@Mercer

Here’s something that I have been waiting for you to acknowledge: if this was a new mutation, shouldn’t it only differ from WT by a single residue?

If it is new then it should have less divergence than the brown genotypes, and it does.

1. Divergence also is dependent on selection.
2. When we address divergence, we have to address both tines of the fork.

So given the mouse germline mutation rate, how long has it been since all of those alleles in the PNAS paper diverged from a common ancestor?

Your hypothesis predicts that the time has to be less than the age of the lava flows.

Why are there four substitutions, all of which change charge? What’s more likely: four simultaneous new mutations or a long history of selection?

@T_aquaticus

It could be one adaptive mutation with the other three hitching a ride.

Could be, but then the divergence time must be far longer if 3 additional neutral missense mutations occurred and became fixed by drift. That time would be more than sufficient to falsify hypothesis A.

So which is more likely?

@Mercer

We’re discussing your unsupportable assumption that it was a new one.

You didn’t say new one.

No, you did:

@T_aquaticus

A new mutation resulting in black fur allowed the mice to take up a new niche in environments dominated by black basalt rocks.

Evolution rarely needs new mutations for rapid selection and fixation, because there is polymorphism in healthy populations.

@Mercer

Who’s to say that the black allele wasn’t the WT in the founding population before it colonized the desert?

If that were the case then I would expect to see the black mc1r allele at both black lave sites. We don’t. We see different genotypes for the same phenotype.

Now we’re getting somewhere. Could any allele be credibly called WT in the founding population?

@T_aquaticus

Just in case you missed it from above:

I didn’t miss it. You’re missing the fact that selection fixing an adaptive substitution, no matter how rapid, has no bearing on the age of that substitution.

[Mercer (John Mercer)]

@T_aquaticus

As a reminder, mutations in just one gene (mc1r) perfectly correlated with the black fur phenotype in the Pinacate population.

As a reminder, I am not disputing that a particular Mc1r allele clearly causes the black fur. You’re trying to claim that this was due to new mutations.

[Mercer (John Mercer)]

@Mercer

Wild populations of both humans and mice are polymorphic at this locus and many others.

That wasn’t the case for the mc1r black allele at the Pinacate site.

Absolutely false; look at Table 1 and Figure 3. That’s the third time you’ve been wrong about the actual data.

@T_aquaticus

Just in case you missed it from above:

Quoting from the paper has nothing to do with my question. Do you really not see from Table 1 that there still is Mc1r polymorphism among the black Pinacate mice?

[Rich_Hampton]

The Wild Experiment That Showed Evolution in Real Time

does this mean that variations in the gene are actually driving color differences in the wild? That question would be very hard to address through more piecemeal experiments. But thanks to Hoekstra’s study, the answer is an unambiguous yes. At the start of the experiment, the delta-Ser mutation was equally common in all six enclosures. After three months, it had become more common in two of the light ones, and rarer in all the dark ones—and the rodents’ fur had shifted accordingly. It clearly provides the variation that natural selection sculpts.

[Mercer (John Mercer)]

The paper is not very clear and appears to have been written by at least two people, at least one of whom correctly calls this a variant allele and talks about changes in allele frequencies, with another who calls it a mutation. Editors often change titles after acceptance.

I don’t know if it’s because it’s late or that the writing is really bad, but I haven’t yet found out for sure whether the deltaSer allele was present in the mice captured on dark backgrounds, which means that the polymorphism was present at the start of the experiment, which means that calling deltaSer a “mutation” is wrong and misleading.

According to the Atlantic article,

At the start of the experiment, the delta-Ser mutation was equally common in all six enclosures [If it is present from the start, it’s a variant and not a mutation!]. After three months, it had become more common in two of the light ones, and rarer in all the dark ones—and the rodents’ fur had shifted accordingly.
[/quote]

I confess that it’s really lame to use the article about the paper to extract data, but I’ll look at it more tomorrow…

[T_aquaticus]

@Mercer

If the mutations occurred after the lava flows, then the time between now and the time of divergence of all of the alleles, both black and beige, shown in the paper must be less than a half-million years. Much less, since the sample is so small and there are almost certainly other polymorphisms that are being missed.

Correct?

The divergence between the beige genotypes would be much older than the divergence point for the black genotypes. The common ancestor of the beige allele would be much older than the common ancestor for the black allele.

@Mercer

Hypothesis C: The founding population was as polymorphic as house mice, with multiple alleles at multiple loci. This allowed them to inhabit multiple niches in the desert, even darker ones like shady, north-facing cliffs. Different black alleles and different black/yellow loci (such as Agouti , the first place the PNAS paper’s authors looked) are likely to have been selected on different flows because of founder effects.

We don’t see the founder effect for the beige allele. Why is that?

@Mercer

So given the mouse germline mutation rate, how long has it been since all of those alleles in the PNAS paper diverged from a common ancestor?

Your hypothesis predicts that the time has to be less than the age of the lava flows.

It has to be less than the beige alleles, and the differences in sequence variation backs it up.

@Mercer

Could be, but then the divergence time must be far longer if 3 additional neutral missense mutations occurred and became fixed by drift.

The neutral mutations would only need to be present in the ancestor in which the single adaptive mutation occurred prior to the adaptation.

@Mercer

Evolution rarely needs new mutations for rapid selection and fixation, because there is polymorphism in healthy populations.

Really? Are the differences in chimps and humans within the variation found in their common ancestor? What of the differences between humans and mice? If we reshuffled the alleles found in the common ancestor of mice and men, could we get both mice and men?

@Mercer

I didn’t miss it. You’re missing the fact that selection fixing an adaptive substitution, no matter how rapid, has no bearing on the age of that substitution.

It’s not how rapid it was, but how recent it was:

Finally, the pattern of nucleotide variation observed among Mc1r alleles from the Pinacate site suggests the recent action of positive selection. Thirteen polymorphic sites are variable among the light haplotypes, whereas only one site is variable among the dark haplotypes (Table 1). The ratio of variant to invariant sites is significantly different between dark and light alleles (1/953 and 13/941, Fisher’s exact test, P < 0.01). The average level of nucleotide diversity among light alleles (π = 0.21%) is >10 times greater than nucleotide diversity among dark alleles (π = 0.01%). The reduced variability seen among the dark Mc1r alleles is the expected pattern if selection has recently fixed an adaptive substitution (26–28).
https://www.pnas.org/content/100/9/5268

[Mercer (John Mercer)]

@T_aquaticus

The divergence between the beige genotypes would be much older than the divergence point for the black genotypes.

You can’t use alleles under selection, much less strong selection, as clocks.

@T_aquaticus

The common ancestor of the beige allele would be much older than the common ancestor for the black allele.

You’re claiming that BOTH the beige and black alleleS (not singular) had a recent common ancestor. How long ago was that, based on the accumulation of synonymous mutations?

@T_aquaticus

We don’t see the founder effect for the beige allele. Why is that?

There are multiple beige alleles. Why is that?

@T_aquaticus

It has to be less than the beige alleles, and the differences in sequence variation backs it up.

I don’t know how a time can be less than a group of alleles. Can you state this more clearly, please?

@T_aquaticus

The neutral mutations would only need to be present in the ancestor in which the single adaptive mutation occurred prior to the adaptation.

Yes. When did the common ancestor of that ancestor and the closest beige allele exist?

@T_aquaticus

Really? Are the differences in chimps and humans within the variation found in their common ancestor?

Many of them, certainly. Are you claiming that none of them were?

Do you really need to resort to such a silly straw man?

@T_aquaticus

It’s not how rapid it was, but how recent it was:

@T_aquaticus

It’s not how rapid it was, but how recent it was:

Finally, the pattern of nucleotide variation observed among Mc1r alleles from the Pinacate site suggests the recent action of positive selection. Thirteen polymorphic sites are variable among the light haplotypes, whereas only one site is variable among the dark haplotypes (Table 1). The ratio of variant to invariant sites is significantly different between dark and light alleles (1/953 and 13/941, Fisher’s exact test, P < 0.01). The average level of nucleotide diversity among light alleles (π = 0.21%) is >10 times greater than nucleotide diversity among dark alleles (π = 0.01%). The reduced variability seen among the dark Mc1r alleles is the expected pattern if selection has recently fixed an adaptive substitution.

“Recently fixed” doesn’t mean “recently mutated,” but I think you know that.