All language subtitles for 57031x_MH_Comp_Inheritance_Patterns_split_05_Codomi-en

MICHAEL HEMANN: OK so corn--

this is, sort of, modern corn on the bottom here--

which we call maize.

And at the top, we have a variety called teosinte.

This is actually generally what was

found in Mexico, and the US, and North America, and Central

America.

It's an endemic plant, and it's considered

to be the precursor of modern corn, right?

So modern corn essentially evolved through our hundreds

of years of de facto genetic engineering

to create what is modern maize, right?

And so if you cross teosinte and you cross maize,

you get an F1 that looks something intermediate

between the two of these, right?

So we would call this roughly a codominant trait,

as was mentioned in a question.

It is sort of a hybrid phenotype.

Prior to Mendel's work, you would

think this is just classic blending--

right-- of two phenotypes to generate

an intermediate phenotype.

OK so say you do this cross, right?

So you cross teosinte to Maize, right?

And you get an F1 generation.

And you cross these F1s together.

Suddenly you see the appearance of particulate genetics, right?

You suddenly see 1 in 500 plants look like teosinte,

and 1 in 500 plants look like Maize.

So you start seeing the appearance

of parental phenotype.

And so the question is, OK, can we

use this information to tell us how many genes actually

distinguish teosinte and Maize?

So--

All right.

So let's think about a single gene.

We'll call it gene A, right?

So gene A-- there is going to be a teosinte allele, which we'll

call A T. These are true breedings, so

homozygous plants, and we'll cross them

with a true breeding, homozygous maize, right?

You get an F1 generation that inherits a single allele

from both of these parents--

that will be heterozygous.

We cross these F1s together to get an F2.

And this F2 can be homozygous for the teosinte allele,

they can be heterozygous, or they

can be homozygous for the maize allele.

And we expect to see these at a 1 to 2 to 1 ratio

of segregating alleles, right?

So 1 in 4 look like maize, 1 in 4 look like teosinte, right?

So, OK, 1 in 4-- so we know it's not a single gene, right?

Because then we'd expect 1/4 of the resulting population

to look like maize or look like teosinte.

So how many genes are actually constituting the difference?

OK well let's think if we had another gene, right?

So we had a B T B T, crossed with the maize alleles, right?

We would expect for the F2--

also 1/4 would look like maize or teosinte at the B allele,

right?

So for two genes A and B, you'd expect approximately 1/4 times

1/4--

or 1/16, right-- would be homozygous for either of these,

right?

So it would be A T A T, B T B T, right?

So if there are two genes, it'd be 1/4 times 1/4.

We know it's actually 500, right-- so 1 in 500.

So basically, we can just calculate how many genes--

so 1/4 to the n-th number of genes equals 1 in 500, right?

So in this case, it is somewhere between 4 and 5, right?

So for four genes, it's 1 in 256.

For five genes, it's 1 in 1,024, right?

So somewhere between, you know, 4 and 5 genes.

That's not a really satisfying answer, is it?

Like, we have 1 in 500, we have somewhere between 4 and 5--

you know, we'd like to have clear answers here.

Why do you think it's not four or five exactly?

Any guesses?

So there's recombination, they may be linked, codominance.

Oh I like this one-- it's only 1 in 500,

the sample size is pretty small.

OK so these are all possibilities, right?

So one of them could be that the person looking at it

isn't very good at recognizing, right?

They look close, and you're just rapidly categorizing

lots of things.

And you're just calling one that when

they're not really identical, they just look kind of like it.

It could be that some of the genes

are actually codominant, that they're

mixing with one another, right?

And it also could be that we just don't

have enough numbers, right?

You know, and you think about it,

if you're actually saying something

is 1 in 500, for that to be different than 1

in 256, versus 1 in 1,024, you need to have appropriate sample

sizes, right?

This is really critical in genetics

because everything in genetics is observation.

We're doing crosses, right?

We're not extrapolating based on first principles,

we're generally looking at numbers and saying,

are numbers consistent with a hypothesis?

And so there, we actually have to start

attaching a statistical test.