How can scientists possibly know ... ?? An open exploration thread

[shernren]

You expect too much. holdon only asks questions. He doesn't answer them.

I am at heart didactic. A real teacher must either have infinite hope or give up immediately.

 

[sfs]

So, you say..... but you don't have any evidence for that. So, it's an assumption; a belief. That's all.

On the contrary, I have a great deal of evidence for that. We can predict theoretically what will happen to genetic variation under the effect of natural selection, and also under randomly accumulating new mutations. Most of the genome looks like the latter, about 5% looks like the result of selection against new deleterious mutations, and a small fraction shows clear evidence for positive selection for new traits.

Nor did I say that.

Then what did you say? What does Hardy-Weinberg say about anything at all that you're discussing?

How can you be so sure they are "all" passed on to offspring?

Because the only ones I care about, and the only ones that we're counting, are the ones that we see in the offspring. We see new mutations all the time in offspring.

 

[sfs]

Really? How do you know?

By reading the papers that describe the research. In the case of the gene M1CR1, variant that contributes to fair skin (and red hair) was completely absent in Africa; the variant that was in Africa was the same as that found in our close relatives. There was also strong genetic evidence that the gene was under purifying selection in Africa, i.e. selection against mutations that would disrupt the gene function. Outside of Africa (where there is much less need for skin pigmentation to protect against the sun), there were multiple variants in the gene.

In the case of the gene SLC24A5, the variant that causes lighter skin was almost entirely restricted to Europe; the other (darker) variant was present at between 93% and 100% in other populations. Once again, the darker variant was the same as that found in other animals. In this case, there was also very strong genetic evidence for positive selection for the new variant within Europe.

 

[holdon]

Yup, holdon: how, in your view, does new genetic material emerge?

First, begging the question: does new genetic material emerge? Or maybe start first with: what is genetic material precisely?

 

[holdon]

You expect too much. holdon only asks questions. He doesn't answer them.

Well, I didn't say "new genetic material" emerges. So, why should I answer the question how it does?

 

[holdon]

On the contrary, I have a great deal of evidence for that. We can predict theoretically what will happen to genetic variation under the effect of natural selection, and also under randomly accumulating new mutations. Most of the genome looks like the latter, about 5% looks like the result of selection against new deleterious mutations, and a small fraction shows clear evidence for positive selection for new traits.

Theoretically.

Then what did you say? What does Hardy-Weinberg say about anything at all that you're discussing?

Because I think most geneticists would agree that Hardy-Weinberg is a good approach to allele frequencies in a population. And that approach shows that purely on a statistical basis, extremes have no chance of making it.

Because the only ones I care about, and the only ones that we're counting, are the ones that we see in the offspring. We see new mutations all the time in offspring.

What exactly do you call "mutations", because I suspect there might be a difference.

 

[holdon]

By reading the papers that describe the research. In the case of the gene M1CR1, variant that contributes to fair skin (and red hair) was completely absent in Africa; the variant that was in Africa was the same as that found in our close relatives. There was also strong genetic evidence that the gene was under purifying selection in Africa, i.e. selection against mutations that would disrupt the gene function. Outside of Africa (where there is much less need for skin pigmentation to protect against the sun), there were multiple variants in the gene.

And has it been established that this M1CR1 is solely responsible for the "fair skin" phenotype occuring in the population? Because you claimed "that the source population (all of sub-Saharan Africa) completely lacks the alleles for blue eyes and pale skin" and concluding that these alleles must be the result of mutations as they are not present in the "source population" which in itself is a BIG assumption. And even if certain alleles do not occur in one part of the species (homo in this case), it doesn't follow that the other part of the species must have mutant alleles, because selection does occur....

 

[sfs]

Theoretically.

Theory tells you what you will see if certain conditions are true. Then you have to look at the empirical evidence to see whether the theory is supported or not. In this case the theory is supported. If you have an alternative theory to offer that would explain the evidence, by all means present it.

Because I think most geneticists would agree that Hardy-Weinberg is a good approach to allele frequencies in a population. And that approach shows that purely on a statistical basis, extremes have no chance of making it.

I'm sorry, but you don't understand Hardy-Weinberg. HW tells you (under idealized assumptions) what frequencies of genotypes to expect in a population, given an allele frequency. It has nothing to do with extremes "making it" or not. (Note: I am a geneticist.)

What exactly do you call "mutations", because I suspect there might be a difference.

A mutation is any change in genetic material. A germ-line mutation is a mutation that can be passed on to future generations, because it is in the reproductive cells..

 

[sfs]

And has it been established that this M1CR1 is solely responsible for the "fair skin" phenotype occuring in the population? Because you claimed "that the source population (all of sub-Saharan Africa) completely lacks the alleles for blue eyes and pale skin" and concluding that these alleles must be the result of mutations as they are not present in the "source population"

I'm concluding that the alleles for fair skin in this gene are the result of mutation, based on the genetic evidence. As it happens, there are something like half a dozen genes implicated in human pigmentation that show evidence for positive selection; this type of evidence only appears when the allele in question starts out on a single chromosome.

which in itself is a BIG assumption.

These variants are not present in the sub-Saharan population; that's just a fact. That the sub-Saharan population is the geographic source of modern humans is not an assumption, but a conclusion drawn from a wide range of evidence.

And even if certain alleles do not occur in one part of the species (homo in this case), it doesn't follow that the other part of the species must have mutant alleles, because selection does occur....

No, but when you know that the first population lives in an environment where these alleles are selected against, as well as being absent (which is true for fair skin alleles in Africa), and when you know that the second population emerged from the first population and moved into a different environment, where the alleles are no longer selected against, you have good reason for thinking the alleles in the second population are indeed the result of mutation.

 

[holdon]

I'm sorry, but you don't understand Hardy-Weinberg. HW tells you (under idealized assumptions) what frequencies of genotypes to expect in a population, given an allele frequency. It has nothing to do with extremes "making it" or not. (Note: I am a geneticist.)

First: you confirmed my point: you also think HW is a good approach. Now, what is the difference between "expect" and "making it"?

A mutation is any change in genetic material. A germ-line mutation is a mutation that can be passed on to future generations, because it is in the reproductive cells..

You said you see "mutations all the time in offspring". I guess you meant what you call "germ-line" mutations. How do you know what you see are mutations? That is: real changes in genetic material, that that genetic material didn't have before?
From a genetic standpoint, since you are the geneticist here, how do explain the incredible amount of hereditary and beneficial mutations that are needed to explain "the origin of species" and at the same time that certain species never changed for an inordinate amount of time, although being subject to the same cosmic radiation and what not as the species that did evolve?

 

[holdon]

I'm concluding that the alleles for fair skin in this gene are the result of mutation, based on the genetic evidence. As it happens, there are something like half a dozen genes implicated in human pigmentation that show evidence for positive selection; this type of evidence only appears when the allele in question starts out on a single chromosome.

It seems that MC1R varies quite easily thoughout the animal kingdom. Nothing proves that the variations of MC1R are "lighter skins" and that the "original" MC1R would be the dark one. It could as well be the other way around.

These variants are not present in the sub-Saharan population; that's just a fact. That the sub-Saharan population is the geographic source of modern humans is not an assumption, but a conclusion drawn from a wide range of evidence.

Yeah, yeah.

No, but when you know that the first population lives in an environment where these alleles are selected against, as well as being absent (which is true for fair skin alleles in Africa), and when you know that the second population emerged from the first population and moved into a different environment, where the alleles are no longer selected against, you have good reason for thinking the alleles in the second population are indeed the result of mutation.

??? Why?

 

[sfs]

First: you confirmed my point: you also think HW is a good approach.

Huh? I think HW is a good approach to looking for experimental errors in genotyping, but it's just about useless for anything else. It says absolutely nothing about evolution: about where variants come from or about how or why they change in frequency.

Now, what is the difference between "expect" and "making it"?

You ask very strange questions. Hardy-Weinberg tells you, for a given allele frequency, how many heterozygotes and how many of each kind of homozygote there will be, in the mean (= "expected value", hence "expect"). "Making it" for you seems to have something to do with whether an allele or a trait survives in the population. The two are unrelated concepts.

You said you see "mutations all the time in offspring". I guess you meant what you call "germ-line" mutations.

Yes, I mean germ-line mutations. You can also see the other kind ("somatic mutation"), which are restricted to an individual, if you care to look for them. Cancers are loaded with them, for example.

How do you know what you see are mutations? That is: real changes in genetic material, that that genetic material didn't have before?

You examine the two parents' genetic material, see what they have for a gene, and examine the offspring's genetic material, and observe something different. Not very complicated, really. The mutations that attract attention, and are likely to be looked at, are those that cause disease, but mutations are happening in the rest of the genome all the time too.

From a genetic standpoint, since you are the geneticist here, how do explain the incredible amount of hereditary and beneficial mutations that are needed to explain "the origin of species" and at the same time that certain species never changed for an inordinate amount of time, although being subject to the same cosmic radiation and what not as the species that did evolve?

Mutations happes -- lots of mutations, all the time. So there is no shortage of raw material for evolutionary change. You will mostly only see substantial visible change in the organism, however, if changing helps the organism survive and reproduce. There is no reason that a well-adapted organism in a stable environment should change much, however long the time period.

 

[sfs]

It seems that MC1R varies quite easily thoughout the animal kingdom.

Certainly. The variation in MC1R causes variation in skin/fur pigmentation, which also vary a lot in the animal kingdom. Nearly hairless apes living in the tropics need a lot of pigmentation for protection from the sun, however, so the gene didn't vary much while we were still all in Africa.

Nothing proves that the variations of MC1R are "lighter skins"

Why do you make statements like this when you don't know whether there is anything that proves this or not? In fact there is quite a lot of evidence that variation in MC1R causes lighter skins and red hair in humans. See here for a summary of it.

and that the "original" MC1R would be the dark one. It could as well be the other way around.

How could it be the other way around? Humans come from Africa. What happened to the variants in the African population in the last few ten thousand years?

Yeah, yeah.

Yeah, really.

??? Why?

Because the lighter-skin variants are strongly selected against in the parent population. That means that those mutations, when they occur, disappear very quickly, since they're bad for the individuals who have them. This is confirmed by the absence of the variants throughout Africa today. So if there weren't in the population that non-Africans descended from, they had to have come from somewhere, and the only only source is new mutation.

This isn't absolutely conclusive, of course: one could imagine scenarios (changing African environment, maybe) such that the obvious explanation is wrong. But there's no evidence for alternative scenarios, and this is very likely the correct explanation.

Here's an idea: why don't you offer something -- an argument, or a piece of evidence, or anything at all -- that suggests some other possibility?

 

[holdon]

Huh? I think HW is a good approach to looking for experimental errors in genotyping, but it's just about useless for anything else. It says absolutely nothing about evolution: about where variants come from or about how or why they change in frequency.


You ask very strange questions. Hardy-Weinberg tells you, for a given allele frequency, how many heterozygotes and how many of each kind of homozygote there will be, in the mean (= "expected value", hence "expect"). "Making it" for you seems to have something to do with whether an allele or a trait survives in the population. The two are unrelated concepts.

HW simply predicts that if certain alleles are rare in P (parent generation), they will be equally rare in subsequent generations, selection not withstanding. In other words they will not be able to make it to the majority of the population, which means you will rarely "see" them.

Mutations happes -- lots of mutations, all the time.

Well to the tune of 1 in a 100 million bases! If you call that "lots" and "all the time", I don't know how you gauge that

 

[holdon]

Why do you make statements like this when you don't know whether there is anything that proves this or not? In fact there is quite a lot of evidence that variation in MC1R causes lighter skins and red hair in humans. See here for a summary of it.

You mean this?: "one might expect mutations associated with red hair to be recessive (see 266300); most of the red-head and fair-skinned individuals in their study were either heterozygous or had no identifiable mutations."

 

[sfs]

HW simply predicts that if certain alleles are rare in P (parent generation), they will be equally rare in subsequent generations, selection not withstanding. In other words they will not be able to make it to the majority of the population, which means you will rarely "see" them

Sorry, wrong. HW says nothing about the action of selection, and only predicts future allele frequency for neutrally evolving infinite-sized populations. For finite populations (i.e. for all real organisms), genetic drift (which has been very well explored mathematically and experimentally) operates. HW is still of interest for finite populations, since it tells what the proportions of different genotypes will be, given the current allele frequency.

[quote
.Well to the tune of 1 in a 100 million bases! If you call that "lots" and "all the time", I don't know how you gauge that[/QUOTE]
1 per 100 million bases (which is probably low by a factor of two) is 60 per person, or 360 billion new mutations in the current population. Looking backward, my copy of the genome has seen ~10 million mutations since it was the same as the chimpanzee genome. I call that a lot.

 

[sfs]

You mean this?: "one might expect mutations associated with red hair to be recessive (see 266300); most of the red-head and fair-skinned individuals in their study were either heterozygous or had no identifiable mutations."

No, I meant all of the studies showing the causal link betwen variations in MC1R and red hair. Did you skip those? Red-headedness is not solely determined by a single gene, but mutations in this gene are clearly one of the causes.

 

[gluadys]

Of course I almost forgot. "The scientific journals" and "scientists".

Quotation marks are unnecessary. Scientific journals are scientific journals, not pseudo-"scientific journals". Ditto with scientists.

Personally, not being a scientist, I rely more on popular science material, but if you are a stickler for evidence, the journals is where it is. Takes a lot of time to research though.

Well documented how? What newly developed species has been well documented? And how?

The key word "speciation" brought up over 7,000 hits on PubMed. That's a lot of documentation.

Of course. Really? Then please explain to me how it is possible that organisms that have not evolved since "millions of years" are still here today exactly the same and unaltered (at least from what we can tell) despite all the environmental selective pressures they must have gone through like all other life forms.

I assume you are referring to "living fossils" like the coelocanth.

1. They are not "exactly alike". Sure they are recognizable as being of the same family as the fossil forms, but they are different species and often a different genus.

2. Mutations alone do not force evolution. Natural selection does. And natural selection can prevent change as well as cause change. When a species is well adapted to its environment, most changes will make it less well adapted. Natural selection always favours the better adaptation, even if it is the status quo.

3. As a source of selective pressures, how much has the deep ocean changed in the last 500 or so million years?

What you call "science" assumes that all life forms must have evolved through an infinite number of mutations, requiring therefore an infinite space of time.....

No, not infinite, but very many.

Or maybe start first with: what is genetic material precisely?

Basically, the DNA found on the chromosomes, the chromosomes themselves, and also some DNA found in mitochondria.

What exactly do you call "mutations", because I suspect there might be a difference.

As sfs said, any change in the genetic material. This includes changes that have no effect on the phenotype at all.

And even if certain alleles do not occur in one part of the species (homo in this case), it doesn't follow that the other part of the species must have mutant alleles, because selection does occur....

If the first population is the source population and does not have these alleles, what cause, other than mutation, exists for the subsequent population to have acquired them?

From a genetic standpoint, since you are the geneticist here, how do explain the incredible amount of hereditary and beneficial mutations that are needed to explain "the origin of species"

Note that there are two ways in which one can have new species. In one case, a single population changes through time but without any second population splitting off from it. In this case we have simply the accumulation of genetic changes, in some cases accelerated by natural selection. The decision to classify the source population and its descendant a hundred generations later as different species is somewhat of a judgment call, as we cannot apply the biological species test of interbreeding. (The closest we get to this scenario is a ring species in which the separation of one population from another via many intermediates occurs geographically instead of chronologically.)

In the second case, the original population splits into two or more groups. At the time of initial separation they are all one species. Separation alone does not make them different species. But since each population is now accumulating a different set of mutations and responding to a different set of selective pressures, each population is modified in different ways. Eventually the application of the biological species test (interbreeding) shows that they are different species.

I mention this to re-iterate again that it is not so much mutations, but natural selection and especially natural selection in populations isolated from each other that brings about new species.

the species that did evolve?

All species evolve all the time, even when they do not exhibit many visible changes or split into separately breeding populations. Evolution is a continuous process, but proceeds at different rates in different species and in different circumstances.

HW simply predicts that if certain alleles are rare in P (parent generation), they will be equally rare in subsequent generations, selection not withstanding.

That is incorrect. Selection will produce a change in the current H-W ratios. There is a formula for determining how rapidly a change in the ratios will occur, connected with the strength of the selective pressure. It involves applying a selection co-efficient to each of the items in the standard H-W formula. This shows how the allele frequencies will change so long as the selection pressure remains constant.

I refer you again to the work I did in this post using selection co-efficients.

http://foru.ms/showpost.php?p=14449626&postcount=5

Here is a student exercise using the pepper moth example, that also shows how selection impacts the frequency by which certain alleles are inherited.

http://cas.bellarmine.edu/tietjen/Laboratories/evolgame.pdf

In other words they will not be able to make it to the majority of the population, which means you will rarely "see" them.

If an allele currently occurring in only 1% of the population is making its way into only 1% more of the population in each generation, it will soon not be rare any more. Eventually, it may be the only allele remaining, i.e. be "fixed" as the species norm.

Well to the tune of 1 in a 100 million bases! If you call that "lots" and "all the time", I don't know how you gauge that

By looking at how many bases you have to work with. 1 in 100 million bases. Well, there are 60 sets of 100 million bases in a single human cell. (i.e. approximately 6 billion bases) At 1 mutation per 100 million bases, that would average out to 60 mutations per cell duplication.

That includes the cell duplication that resulted in the germ line cell that is inherited by the next generation.

And that figure is somewhat low by the actual estimates I have seen.

If 60 mutations in every zygote is not "lots" and "all the time" what is?

 

[holdon]

No, I meant all of the studies showing the causal link betwen variations in MC1R and red hair. Did you skip those? Red-headedness is not solely determined by a single gene, but mutations in this gene are clearly one of the causes.

I agree that variations in MC1R are among the causes for red hair. But it doesn't mean that mutations cause this, per your the puzzling quote from your article.

 

[holdon]

Sorry, wrong. HW says nothing about the action of selection

But I didn't say that it did.

and only predicts future allele frequency for neutrally evolving infinite-sized populations. For finite populations (i.e. for all real organisms), genetic drift (which has been very well explored mathematically and experimentally) operates. HW is still of interest for finite populations, since it tells what the proportions of different genotypes will be, given the current allele frequency.

Of course genetic drift must now explain everything. But it doesn't, because it assumes.

1 per 100 million bases (which is probably low by a factor of two) is 60 per person, or 360 billion new mutations in the current population. .

Actually, the number should be increased by a factor or 2, yes. Anyway, who knows what the number really is? Nor does it account for the even lower likelihood of persistance: hereditary.Nor for supposed repeated "beneficial" ones.

Looking backward, my copy of the genome has seen ~10 million mutations since it was the same as the chimpanzee genome.

That explains it.