Introduction to Genetics and Evolution is a college-level class being offered simultaneously to new students at Duke University. The course gives interested people a very basic overview of some principles behind these very fundamental areas of biology. We often hear about new "genome sequences," commercial kits that can tell you about your ancestry (including pre-human) from your DNA or disease predispositions, debates about the truth of evolution, why animals behave the way they do, and how people found "genetic evidence for natural selection." This course provides the basic biology you need to understand all of these issues better, tries to clarify some misconceptions, and tries to prepare students for future, more advanced coursework in Biology (and especially evolutionary genetics). No prior coursework is assumed.
Searching for Natural Selection: McDonald-Kreitman Test (S)
Loading...
From the course by Duke University
Introduction to Genetics and Evolution
503 оценки
From the lesson
Molecular Evolution
This advanced module explains why sexual reproduction (involving recombination) is evolutionarily advantageous, and discusses how the analysis of DNA sequences can be used to understand the evolutionary forces acting on populations or species, either in general or at specific genes.
Meet the Instructors
Dr. Mohamed NoorEarl D. McLean Professor and Chair,
Biology
Hello, and welcome back to introduction to genetics and evolution.
In the previous video we talked about looking for evidence for
natural selection at individual genes.
And try to focus on differences say between humans and chimps.
And infer whether these differences that we see may have been the result in
part from the action of natural selection versus genetic drift.
dN/dS test is useful in this regard.
This is looking at the number of non-synonymous changes,
per non-synonymous site, divided by synonymous changes per synonymous site.
The advantage of scaling the synonymous changes,
is these are expected to be neutral.
Whereas the non-synonymous changes may be advantageous, may be neutral,
may be detrimental.
The problem though is the dN/dS can sometimes mix
up constraint with adaptive evolution and just call it neutral.
So this is a little bit of a challenge and dN/dS is generally believed to be fairly
conservative when looking for adaptive amino-acid.
When I say conservative I mean there's too many false negatives.
There's too many times when you fail to see positive selection that actually has
potentially operated in the evolutionary history of a particular gene.
Now we can leverage some aspect of neutral theory that we haven't talked about yet.
And that is that we predict the ratio of nonsynonymous to synonymous changes
should be roughly constant through time, right?
The ratio that you see among individuals within species
should be roughly similar to the ratio you see between species.
You may be wondering, why should this be constant through time?
Well, there's an assumption about the neutral theory that most non-deleterious,
non-synonymous mutations are neutral.
Basically, those mutations that you see within a species that are just segregating
around, they are at some frequency, those tend to be neutral,
because the bad ones are eliminated, and the advantageous ones spread.
So you tend to see a lot of these.
So these non-deleterious,
nonsynonymous mutations will behave like synonymous mutations.
And they'll rise at some sort of constant rate.
And they'll fix with about the same probability as a synonymous change.
This idea of the allele frequency should predict 1/(2N).
Now two fellows, John McDonald and Marty Kreitman, suggested this test,
which has subsequently been named after them.
They didn't call it the McDonald-Kreitman test, they just named it a test.
Now they proposed that you can test for a selection by seeing if this ratio is
constant within species and between species.
The importance here is this within species is looking at the present day,
we're taking a snapshot of a population right now.
What do we see in terms of nonsynonymous variation, synonymous variation?
The between species indicates historical changes,
what's happened over the evolutionary history.
So what they do is, they align sets of sequences, and
identify their variable nucleotide sites.
They basically just have this four cell test,
where they identify each difference as a nonsynonymous difference within species,
a synonymous difference within species, a nonsynonymous difference between species,
or a synonymous difference.
So let's try this, that was a lot of words, let's try it in practice.
So here we have an example, let's say that we have four individuals in species one,
and one individual in species two.
For the sake of argument, let's say species one are humans and
species two are chimps.
Now here's our code on table again and what we're looking for
are these two parallel categories.
First, what is fixed between, so these are differences between species one and
species two.
We have polymorphism within, so
basically what is variable within, in this case species one, and
then we can also categorize any of these differences as being nonsynonymous or
amino acid changing, or synonymous where they don't change.
So what we do is we just walk through.
In this case we're not looking for a site, we're actually just looking for overall.
So where are the differences between species here?
Well, here is a difference, within or between species.
So there's a difference here between species,
this is within species, this is between species, this is within species.
This is within species.
So we have an amalgam here, we have some within species difference, and
some between species difference, okay?
Now let's walk through them.
So this first site here, this is a non-synonymous change as you can see here,
CT anything to ATG, which is Methionine.
That is a non-synonymous difference and
this is a non-synonymous difference between species.
So that goes in this cell.
We skip this one because we're not going per site,
we're just gonna skip the next one and go to the third nucleotide site.
In this case we have a difference within species, but they all are making Leucine.
So it's polymorphic within, all right?
And synonymous.
So let's jump now over a little bit.
This one right here is a difference between species, but again AC anything
is Threonine so this is a synonymous but between species difference.
Jump over a little bit this is a within species difference but
second position, second position is always non synonymous.
So it should be nonsynonymous within and
the last one over here, ACC or ACA those are both threornines.
So there should be another polymorphic within species because the variation is
just in species one, species two has one of the ones from species one and
is synonymous.
So we basically tally these things up and we have these four numbers that come out,
we'll just call them A, B, C, and D for the sake of argument.
If all nonsynonymous differences are neutral,
then we expect A/C,
this ratio you see between species, is essentially reflecting the historical,
should be equal to the ratio B/D, what you see within species, or the present day.
In contrast if some, or especially if there's a lot of non-synonomous
differences between species that were advantageous and
selected, like let's say something helped us humans
be able to digest things that are unique to our diet better.
Then this ratio should be greater than this ratio, that the AC > B/D.
What's happened then is these difference will have spread within humans.
They're changing the amino acid ratio.
Since they spread within humans, they're now a between-species difference.
They're not something that's neutral that's just kind of hanging out
in the population, okay.
So let's look at a couple of examples.
Here's some actual data from a few genes.
Here's SEMP1, that enables cellular survival under low oxygen.
This is a case where we see a large difference in terms of nonsynomis or
synomis between species relative to within species.
This would be a good candidate for positive selection.
There's been some sort of adaptive difference between humans and chimps.
In contrast, this one CIAS1, which affects an auto-inflammatory response,
that one is pretty similar between the two.
So, this is a case where we can't reject neutrality.
The ratio here, five tenths, is equal to one half.
Now what about the opposite?
This is what we haven't talked about.
What happens when, instead of being greater between species,
or equal, what happens if it's the other way?
So here's the gene AGT.
The mutations in AGT affect hypertension, associated with coronary artery disease.
What happens if you have a high number of non-synonymous differences among humans?
A higher ration non-synonymous difference among humans,
than between humans and chimps?
What do you think that might mean?
So now we have to think about what kind of variation would do this.
So what kind of variation persists for a long time, but doesn't fix?
Well the answer in this case, Is unfit recessive alleles.
So let me show you a picture, and this is a picture you saw before.
Let's say over here on the x axis is time and generations.
On the y axis is allele frequency.
Now, let's imagine hypothetically that you had an allele that was actually bad, but
it was recessive.
What you might see is you might see something that looks like this.
We saw this, actually, in previous videos,
when we talked about the action on the actual selection, that it's
actually kind of hard to completely get rid of these recessive bad alleles.
These bad alleles can stick around for a long time, because what's happening here,
let's say, for example, the bad one is aa.
And then the good ones are AA and Aa, right.
What happens is it gets very rare,
almost all of the little as are present in the heterozygous form.
So they're basically masked and selection doesn't get rid of them very well.
Only the very small number of aa are actually being eliminated by
natural selection.
So the better dominant allele never quite gets to fixation.
Now a lot of diseases, as I mentioned way back when, tend to be recessive.
And as a result these mutations causing disease will stick around
within populations but never fix, and they eventually do get lost, but
it takes a very long time for them to get lost.
So this is a variation that you'll see that will stick around within the species
and it effects the within the species numbers,
but does not effect the between species numbers.
And these diseases will tend to be associated with non synonymous
bad variations, so this is what causes that.
Essentially, here's the ratios again we talked about,
between species nonsynonymous over synonymous versus within species
nonsynonymous over synonymous.
That if all nonsynonymous differences were neutral we
expect these ratios to be similar.
If some of these nonsynonymous differences were advantageously selected then
we expect this ratio, A/C > B/D.
Right, because the advantageous ones would spread,
they would come between species non-synonymous variation.
In contrast, if you have maladaptive non-synonymous differences,
they will persist within species and
never be converted to between the species differences.
In that case A/C < B/D.
The that's what will happen if you have diseases lingering around.
So looking at the McDonald Kreitman test we have this A/C = B/D.
That's a case where you can't reject neutrality.
Again, it doesn't mean it's neutral, it can be a mix, but you can't reject it.
If A/C is greater than B/D,
then we have increased non synonymous changes between species.
We have more non synonymous changes than we expected between species relative to
within And that is indicative of positive selection or advantageous variation.
So you're running.
In contrast, if A/C is less than B/D then you have decreased nonsynonymous changes
between species.
What that means is you have some variation that's lingering within species and
doesn't quite get fixed.
That is often the result of negative selection.
There are other possibilities, things like overdominance that we talked about earlier
will also do this same sort of pattern.
But it's thought that over dominance is not especially common, so
a lot of this is going to be in the boat of negatives selection or
bad variations sticking around.
So what happens when we look at the genome?
I showed you a couple of examples.
This is looking at a lot of genes studied across or
the difference between humans and chimps.
The blue marked genes are ones that
have McDonald-Kreitman tests results indicative of positive selection.
The red ones are indicative of negative selection.
You notice there's a lot more red out there than blue.
But again, we know that most mutations are bad.
We know that most bad mutations are recessive, therefore,
it's gonna make it so they're just kind of lingering around.
So instead of just looking at this figure, let me give you some of the numbers.
The positively selected genes from that particular study I was just showing you,
there are about 304.
Again, what are they associated with?
The same sorts of things as the high dN/dS values, immunity protein genes,
gamete formation ones, also some for sensory perception.
In terms of negatively selected ones, many of these are involved in cytoskeleton
formation and they're often associated with diseases like, muscular dystrophy,
cardiovascular disease, etc.
So let me give you one to try here, so you can have a little bit of practice within
the video before you even try any of the practice problems.
So this is a hypothetical sequence for arsD.
This is something that controls arsenic resistance, okay?
So I want you to go through here and calculate and
interpret the McDonald-Kreitman test.
So, you may want to pause the video before you actually get to the in-video quiz.
Yep. I hope that wasn't too difficult for you.
Let's go ahead and take a look.
So what I'm gonna do is I'm gonna highlight things in blue and red first.
So the blue ones here are indicating differences between species, so
this is comparing humans to chimps.
The blue ones are indicating differences between species.
The red are indicating sites that are variable within species,
to make it a little bit easier to see.
So let's look at those.
We're looking at between species.
We have one non-synonymous difference, that's this one right here,
this is non-synonymous.
And we have two synonymous differences between species.
All right, so this is a synonymous difference here, and
synonymous difference here.
When looking within species, what do we see?
We have two nonsynonymous.
This one is nonsynonymous.
Remember, always the second side's nonsynonymous.
This one's also nonsynonymous.
And this one right here is synonymous.
So one synonymous within so our ratio's 1:2 versus 2:1.
So again, which one is bigger?
And of course we see it's bigger in the within species category,
that this is a result in the direction of negative selection.
That there is potential bad variation sticking around within species and
not being eliminated.
So let's recap real quickly, here's the McDonald-Kreitman test.
Again if A/C, A being the between species nonsynonymous divided by between
species synonymous for C is equal to B/D, the same for within species,
you can't reject neutrality, does not mean that it is neutral.
If it greater between species than within species, then you
have increased nonsynonymous changes, that interesting positive selection, right.
In you have A/C < B/D, you have decreased,
you have been holding back the nonsynonymous changes.
They're staying within species and not converting to differences between them.
That is negative selection.
Now I've shown the dN/dS test and
I've shown you the McDonald-Kreitman test, there are other metrics out there.
There are many that also exist that we won't use in class,
the one popular one is referred to as Tajima's D.
That uses the frequencies of alleles to test for recent selection or
recent changes in population size in example.
But importantly these are implemented to understand what's going on.
There's two general approaches for study.
One is to scan the whole genome using these metrics and look for
possible selection and then try to interpret it.
All right, so that's like the picture I showed you.
They're showing all the red and blue labelled genes.
That's where we look at everything and try to interpret how much
positive selection versus negative selection in the genome.
The other is the look at candidate genes,
to pick out the subset of genes that you know effect brain size or
effect speech and look for the signatures of the selection of those.
Or maybe even individual genes.
Now, both of these are active areas of research.
And both of them have given us some clues as to what's been important in our
divergence or the divergence of any other species that have been studied.
So closing out, what makes humans special?
Well, these DNA sequence-based tests have shown that, again, hundreds of genes have
undergone selection to change, and a lot of that seems to be, when I say selection
to change, I'm referring to this positive selection, this adaptive variance.
More genes are under some selection to stay the same.
Nonsynonymous changes are bad, and therefore they're held back.
They don't ever become differences between species, but
may linger as differences within species.
Therefore some of these bad alleles are sticking around, and this also makes it so
a lot of our sequence is a lot more similar chimpanzee,
than if we were evolving purely neutrally.
But importantly, and let me just close with this,
there's no single gene change that made us what we are today.
Yet, we're not as different from chimpanzees,
genetically, as we might like to think.
Well, I hope you found this interesting.
Thank you for joining.
Ознакомьтесь с нашим каталогом
Присоединяйтесь бесплатно и получайте персонализированные рекомендации, обновления и предложения.
Coursera делает лучшее в мире образование доступным каждому, предлагая онлайн-курсы от ведущих университетов и организаций.
© Coursera Inc., 2018 Все права защищены.
