Tuesday, June 2, 2015

Imagine no mutation.

Imagine no mutation.
I wonder if you can.
No means for new variation.
A creationist's view of Man.
Imagine all the people
learning it this way.

Boo-hoo, ewww.

I found out this semester that although most of my students in "Human Origins" had learned about natural selection, that relatively few of them had learned about mutation. 

Huh? Are people still stuck on Darwin? Apparently. 

Mutation requires a little bit of genetics to understand, but c'mon! Natural selection makes no sense without it. None.

And what's more, natural selection cannot be the all-powerful force that popular culture might have us believe when mutation is accounted for. Evolution is so much more fascinating with even a basic understanding of mutation. In this molecular-clock, whole-genome-sequencing era, how can anyone teaching and learning evolution not have at least a basic understanding of chance, perpetual accumulation of mutations?

We've all, each and every one of us, got many de novo mutations. So our genomes are distinct from our parents' and siblings', not just because of genetic recombination of our grandparents' genomes when our parents' eggs and sperms were built, but because of unique mutations that occurred while making those germ cells or in the early stages after their union.

That this constant change in lineages is occurring with each reproductive event is proof that natural selection is a largely tolerant process, perpetually allowing perpetual evolution by mutation.

(And, maybe natural selection has much less, and mutation has much more, to do with speciation than many have assumed.)

What happens to each and every one of our unique mutations, whether or not they live on in our offspring, whether or not they play a role in adaptation, depends on quite a bit of luck, partly because of the "Law of Segregation":

A Mutation's Future
Click to enlarge. Email me for original file that you can modify: holly_dunsworth@uri.edu

For a description of evolution that incorporates perpetual mutation, check out "Evolution is the only natural explanation. And it's all we need."


Ken Weiss said...

One problem is that many, even biologists and anthropologists, are sloppy about this. They argue that in any population there is enough standing variation, scrambled around in the population by recombination, to respond to changes in the environment. They then effectively equate that, under selection, to 'evolution'. They would all recognize that mutation must occur, but argue that the population doesn't have to wait around for mutations to occur in order for it to adapt to new circumstances. So it's sloppy professional semantics. The issues are elusive, and one good source for the view that mutations are more important than selection is Masatoshi Nei's book 'Mutation-driven evolution'. He goes over most of the viewpoints and subtleties.

There are some confusions in all this. Mutations include functionless DNA changes and these change frequency by drift, and this is a form of 'evolution'. But it's not 'adaptive' and many biologists don't seem to want to count that as evolution. And there is speciation, which is what Darwin wanted to explain by selection, while Nei and others argue that mutation is required for species barriers (mainly, mating incompatibility) to arise, that is, that selection alone won't do the job. Darwin did some very professional-level hand-waving about this! So for some, evolution is about species formation, while for others it's about the generation of variation and its distribution.

Your post gets right at some of the important issues related to selection and mutation, and should be helpful in teaching. But the resesarchers themselves should be clearer about what they mean by 'evolution', 'speciation', and so on.

Holly Dunsworth said...

Thanks for this comment Ken.

Surely they know something about mutation, but I think what my students are telling is me that they thought mutations were rare and terrible, mostly, and only rarely are wonderful, but when they are, natural selection will pounce. This is not only the popular misconception but it's how actual scholars describe things in academic publications, like in books published by university presses! I have examples but I don't want to shame the authors, publicly.

I don't think my students would word the latter half that way, about how "natural selection will pounce" on good mutations, and I know they know that mutations are (as far as anyone knows) random, but I know they're not thinking these things through properly. I pose a series of questions in the form of scenarios where large finch beaks "will evolve" if all the small seeds are eaten, or that "every" male gorilla in the next generation "will be" larger than the previous generation's males if selection for large body size is occurring. That sort of thing. So many fall for this and agree! And others just learn quickly to see a prediction, hear some devil/angel their shoulder whispering that you can't predict many details of future evolution except that it will occur, and they will disagree but they really don't know why they shouldn't fall for it other than that. And that's what I'm trying to help them get at. Mutation and population thinking and chance inheritance and chance expression and environmental change and environmental interactions and multiple genes for a single measurable trait and developmental influences. Oh my....

Holly Dunsworth said...

I just changed the heading on the figure to clarify it.

Ken Weiss said...

The subject is fraught with subtlety, and even scientists (much less text books) miss or ignore them. And then we hear that students (naive poor babes that they are) can't understand the subtlety so we'll just teach them some 'simplified' version (that is, some ideological version).

The idea that most mutations were clearly harmful was known to be wrong more than a century ago, even before anyone knew about DNA. And what does 'random' mean, and how would one test it? (random relative to (a) function, (b) local sequence details, (c) some other criterion?). When these sorts of things, which are not so complex that students (poor things!) couldn't grasp them, then of course they will think, as you say, 'every' male gorilla will.....

And all of this before any of the self-promoting advocates, political or religious demagogues get hold of the oversmplified versions of what's known.

So your effort to try, at least, to give people a clear picture is laudatory--and needed.

Holly Dunsworth said...

I forgot to include in my figure that the mutation itself can mutate.
These ideas are always mutating aren't they?

Ken Weiss said...

Good point. And whether a mutation is 'good' or 'bad' in relation to the environment (i.e., natural selection) changes as the context changes. So no absolutes! Every truth like any spot in a genome, can mutate, again and again.

Anonymous said...

Nice graphic.

(The following is a little facetious.) There are two kinds of evolutionary biologists: those who think that selectionist explanations are unsophisticated, and those who think that nonselective processes are irrelevant.

It is a bit amusing that 73 years after Huxley's Evolution: The Modern Synthesis the relative roles of selection, mutation, and drift are still controversial and that camps of adaptationists, mutationists, and structuralists with intellectual lineages traceable to deep in the pre-Synthesis era are still around.

Evolution is a trickster, consternating those who would try to understand it.

Peter said...

I think "unless it arises again" in the green box is a bit misleading, depending what organism you're talking about. For humans, there's no "unless" about it: the mutations are pretty much guaranteed to arise again.

Per Larry Moran's calculations here, every nucleotide in the human genome will on average acquire a mutation in someone every 20 years or so - a mere blink in evolutionary time. At the very most, you'll have to wait a few generations for a given mutation to come round again. Many of our domestic animals will be present in similar numbers. Insect species, could be much higher. Tigers and whales, much lower.


I imagine there's some interesting work to be done (probably has already been done?) looking at evolutionary processes in terms of whether a given species is mutation-limited. That is, whether the population size is large enough that all nearby "single-step" variants are continuously present in the gene pool, or whether the population has dropped low enough that the new mutation rate is negligible and the population is only re-shuffling existing genetic variation.

That will determine whether the population is sampling its entire local fitness neighborhood, or only a small subset of it. In my mind, this distinction has the feeling of a phase transition, but I'm not enough of a mathematician to say what that implies.

Ken Weiss said...

It seems that there is no attention, grants, papers, news stories or promotions to gain by noting that all these factors are involved in a context-specific mix. No, we want headline claims.

Holly Dunsworth said...

Peter, thanks so much for your comment! These are great things, many new, for me to think about. And surely "unless it arises again" also depends on the kind of mutation we're talking about!

Making the figure was tougher than it might seem, not just because of my intellectual limitations, but also because, for example, "mutation" is left undefined. Does a repeat of a specific point mutation really constitute a repeat of that mutation if the rest of the genome is now different because of recombination and other mutations? And what about copy number variation? What about deletion? What about jumps within and between genomes? Insertions from viruses? And there are other phenomena that fall under "mutation." So although I did lead readers to assume the figure's talking about errors in meiosis or early zygote cell proliferation, that does limit the mutational possibilities I suppose, but it still does not limit them only to point mutations.

Ken Weiss said...

With a mutation rate of about 1.0E-8, and a genome size of 3.1E9 and we're diploid, and a human population size about 7E9, each point mutation occurs many times a year in someone--reinforcing Peter's first point. And this is for single-copy nucleotide changes; if you actually include the mutation rates at polynucleotide runs, microsatellites, telomers, centromeres, and gene duplication and other copy number variants etc., the rates are much higher. What is unique and basically doesn't recur, is the same mutation in the same haplotype or genotype context.

Anonymous said...

Isn't the relative importance of mutation vs selection a function not only of population size but the effect of the mutation?

1) Mutations to noncoding DNA or that are silent (result in same amino acid) and hence invisible to selection, responsible for molecular clock, 2) Mutations of extremely slight effect on protein function have important role in reproductive isolation, create clock-like speciation (that new paper, every 2 million years or so), 3) mutations of small effect are more visible to selection, tweak developmental networks, result in maddeningly weak but significant phenotypic correlations, reappear frequently enough in large populations that provide search space, often functionally replaceable with other mutations 4) mutations of moderate effect are very visible to (mostly negative) selection, more obvious to researchers but less of the bulk of genetic-developmental networks than 3) 5) mutations of very large effect - homeotic, meristic, topological etc. effects - create higher order clades (saltation), extremely rare, overwhelmingly negative but those that are subject to positive selection are extremely important.

From 1-5 these get increasingly rare. (Of course the 1-5 formulation is a bit of a fiction since its a continuum.) 3 and 4 give selection a strong role in shaping organisms by exploring search space of (relatively) common and recurring mutations with significant (though often subtle) functional impact. 1 & 2 lock out selection for the most part, create clock-like divergence and speciation. The rarity and importance of 5 gives selection little to do except wait around and certify (to use grossly anthropomophic language) the 'arrival of the fittest'. So a graph with 'effect size of mutation' in the X axis and 'role of selection' in the Y would look like an upside-down U.

Ken Weiss said...

I would agree with your opening line, but your 1-5 points could be highly debated. More important, to me anyway, is that unfortunately we don't have a rigorous theory about the mix of 1-5 and variants can change categories, too, depending on circumstances. Single mutations can generate species if they impair fertility, most would argue that large mutations are most likely strongly selected against, and I think it's a mistake to say they are ones species have to wait around for, since experimental evidence strongly shows that most traits can respond to selection. However, that is consistent as well with Wagner's 'Arrival' argument. One can think of homeotic changes as large, or as rather small when it comes to questions of new 'innovation', and this gets into issues of gene duplication and its role.

Anyway, to me personally all the points are relevant but too complex for simple generations.....and, in particular too complex for simple response to a commentary!