0:00

Hello, and welcome back to Introductions to Genetics and Evolution.

Â Now what we've been trying to do in the last couple of videos

Â is quantify when we're looking at variation in some traits,

Â such as human height, how much of the variation you see is genetic, and

Â how much of the variation you see is environmental.

Â Now, in the previous video we estimated heritability, or that fraction

Â of the overall phenotypic variation that is genetic using an F2 cross.

Â Here's the formula by the way for heritability.

Â Again, that fraction of the overall phenotypic variation is genetic.

Â 0:34

We estimate using an F2 cross.

Â And we start off with tall parents, that are genetically identical to each other,

Â so the genetic variance here is equal to 0.

Â We cross into short parents, they're all genetically identical to each other.

Â So within this box, Janet grants a zero.

Â We got intermediates, the offspring here from them.

Â The genetic variation in these is also zero,

Â they may be heterozygous for a lot of low sine, but they're all heterozygous for

Â the same low sine exactly the same way.

Â Okay, so this is the F1s.

Â From these, we got the F2s, from the F2s,

Â we see quite a bit more spread than we saw in the F1s or in the parents.

Â The reason we see more spread is because we've added this.

Â Instead of Vp just being Ve, Vp here is Ve plus Vg,

Â we've added genetic variance here.

Â 1:32

This is not very useful in the context of something like, for example, human height.

Â Because we don't have a bunch of people who are all genetically identical,

Â they are tall.

Â And we can cross to a bunch of people that are all genetically identical,

Â they were short, and look at the offspring and

Â their grandkids, and obviously it's not feasible.

Â Instead, another way to estimate heritability is to use what we refer to as

Â parent offspring correlation.

Â 1:56

If we assume that all variation is genetic, and

Â let's assume there's no dominance in this case,

Â then any individual should have exact average phenotype of the two parents.

Â So, here's a picture of me.

Â Here's a picture of my wife.

Â If you put us together into some online form,

Â it predicted this is what our kid should look like.

Â This is some sort of intermediate between me and my wife.

Â In contrast, my actual son looks like this.

Â There's a picture of him.

Â Close to, but not exactly identical to this exact average.

Â Now oh yeah, he's not usually quite that tan but

Â he was out, out in the sun a little bit much that summer.

Â 2:35

Now if we look at lots of individuals, we can assess how well

Â the average of the parents' traits predicts the average of the offsprings.

Â And then we can use that.

Â We can look at the strength of the correlation or in this case,

Â the slope of the line to estimate heritability.

Â So let's imagine a hypothetical thing.

Â Let's take an average height of two parents, and

Â let's say the average height of two parents was 6 foot 0.

Â So let's take another average height of parents and

Â say their average height was 5 foot 0.

Â If the kids on average have exactly the same height as the parents

Â the average of the kids matches the average of the parents.

Â That slope is one; therefore the heritability is one.

Â That is a strong genetic component.

Â In contrast let's imagine that you see something looking.

Â Be more like this.

Â 3:24

So you do, you look at the correlation between these two,

Â between the height of the parents and the height of the kids, and

Â you see that in general, there's no particular prediction whatsoever.

Â The slope of this line is zero, therefore there is no prediction,

Â from the average of the parents, to the average height of the kids.

Â This would suggest there's no genetic.

Â We can use this to assess heritability.

Â So I have here four figures.

Â In Figure A, this is a case where heritability would be approximately zero.

Â Where you have short versus tall average for the parents,

Â short versus tall average for the offspring.

Â You can see the slope there is basically zero.

Â 3:57

These you see some relationship.

Â This would be about a case with heritability is about .5.

Â See here you have a very strong relationship.

Â There the heritability is very very close to one.

Â In this case the height of parents, the average of the two parents,

Â very well predicts the average height of the offspring.

Â And here's some real data over here.

Â This is looking at students and their parents.

Â This is from the Evolutionary Analysis textbook.

Â You see the midparent height is very close to the mid offspring height.

Â The slope there is 0.84,

Â suggesting a very stong genetic component to human height in this case.

Â 4:28

Now that's interesting, but I have to remind you why this matters.

Â Again, there is medical relevance to this.

Â Let's say your parents get gall stones.

Â Is it worth it for you to alter your diet?

Â Right? Or is this predominantly genetic?

Â Do you as a doctor want to tell somebody, no, give up eating meat completely?

Â When in fact giving up eating meat may have no effect whatsoever on their

Â likelihood of having some sort of disease.

Â No, you need to be able to identify something as having a strong genetic

Â component versus a very weak genetic component and

Â more of an environmental effect, so you can practice medicine properly.

Â Similarly, let's say you want to breed a friendlier guinea pig.

Â How much will selective breeding matter?

Â Or, is friendliness just how much you handle the guinea pig,

Â something along those lines.

Â Unfortunately, things aren't always as straightforward as I said.

Â Now everything we've been talking about, ignores the fact that parents and

Â offspring may share some environmental factors, as well as genetic factors.

Â There's probably a correlation in food availability and

Â other characteristics between parents and offspring.

Â And what this does is this biases upward the estimate of heritability.

Â We are assuming anything that's shared in this way is from genetics but, in fact,

Â there may be some environmental affect that's shared,

Â especially in this sort of parent offspring correlation that I talked about.

Â There are other assumptions as well.

Â 5:42

First that we're assume the environment is constant, such as for

Â example that F1, F2 cross we looked at.

Â We assumed that Ve was exactly the same in the F1 as in the F2.

Â But in fact the environment is probably not constant.

Â Estimates of heritability may also be very different

Â if you go to different places because Ve may be very different.

Â This is also true at different time.

Â 6:02

And finally, the amount of genetic variation is also not constant

Â in different families or populations.

Â So let's say for

Â example you study in Australia the heritability of human height.

Â That result may not apply if you went to India, if you went to Japan,

Â things like that, because the amount of genetic variation is going to be

Â different, the amount of environmental variation may also be different.

Â But, despite all this, it's not perfect but it's a starting point.

Â And it's still useful under particular circumstances.

Â 6:31

So next class, we'll actually go into another means for

Â estimating heritability and that is what's referred to as the breeder's equation.

Â We'll also talk a fair bit about population growth,

Â that will be a little bit of a side.

Â Thank you and I hope you enjoyed this.

Â