All language subtitles for 77031x_MH_Tetrad_Analysis_06_Applications_Equations-en

MICHAEL HEMANN: So let's put in some numbers here.

So we have just that formula again, 100 times

T plus 6NPD over 2E.

So say you have 75 PDs, you have 20 T's, and you have 5 NPDs.

So here we have your distance in centimorgans

equals 100 times the number of T's plus 6 times the number

of NPDs over 2 times the total number,

which is 100, which equals 50 over 200, or 25 centimorgans.

So think about this.

If you just counted random gametes,

so if you just counted random gametes, you would have,

obviously, T's, and NPDs, and PDs, but you're

just going to count just single cells without regard.

So you'd expect for every T you have,

so you have 20 T's, you'd have two recombinants within those

T's.

So that would be the T class.

Plus you'd have four recombinants for every NPD

that you had.

So you're just counting these randomly.

So T's and NPDs.

So if you looked at 100 times 40, so 20 times 2

plus 5 times 4 over the total number of gametes

that you have, so it's 100 times 4,

it would give you 15 centimorgans.

If, essentially, we're not taking the double recombinants

into account, we're systematically underestimating

genetic distances.

And so this ability to look at tetrad types

allows you to essentially infer the existence

of double crossovers based on the number of NPDs,

so the number of double crossovers that we don't see.

The double crossovers that are tetra types,

or the double crossovers that come from--

that look like parental ditypes you remember here.

So for example, in this case where

we have a first crossover of 2, 3,

and a second crossover of 2, 3, this

resulted in a parental ditype.

But in reality, there were two recombination events here.

We're just not seeing them, because

the second recombination shifted the initial recombination

event back.

So what the NPD stuff does is it allows us to look at--

infer the existence of these double crossover events.

Now importantly, this formula can be applied

if there are no NPD events.

So if there are no NPD events, it's just 100 times

T plus 6 times 0 over 2E, which equals

T over 2E, which is just the formula that we had before

for a tightly linked gene.

So you can include it, even as a general formula.

And you can also think OK, well, what

happens if you had totally unlinked genes?

So we would expect this ratio of one to four to one

if they're unlinked.

So 10PD to 40T to 10NPD.

So what happens if we actually put-- so unlinked numbers

into this equation?

So say we have 100 times the number

of T's, so in this case, say, 40 T's, plus the number of NPDs

times 6 over the total, or 2E, which would be 2 times 60.

I'm just putting in these numbers here into that formula.

We would get 100 times 100 over 120, which is something

like, I don't know.

What is it, 83 centimorgans?

So if they're unlinked, it's giving us

a number that's greater than 50 centimorgans, which suggests

that they're in fact unlinked.

So it allows us to generally put in numbers, regardless

whether they are linked or unlinked,

to get some kind of genetic distance.

And again, if you're going to try to exclude some hypothesis

or take another into account, your numbers are going to--

the numbers that you have, the number

of tetrads that you analyze are going to matter, absolutely.

So numbers do matter here in making a statistical inference

based on this data.

But with the data in general, or what the approach generally

allows you to do is remember, we had this issue

before comparing genetic distance and physical distance

where essentially in the other systems,

in the Drosophila system and also in mammalian systems,

this is really breaking down early on.

So basically at 50 centimorgans, there's

really no ability to discern any distance whatsoever, but really

breaks down significantly before them.

With yeast and tetrad analysis, we

can actually look at accurate interactions,

maybe up until 40 centimorgans or so.

Once you start getting to greater genetic distances,

the recombination rates, or double recombination rates

really obscure this connection between genetic distance

and physical distance.

So this tetrad analysis again provides much more accurate

measures of genetic distance.

The real truth here is that that 50 centimorgan distance

is the distance at which recombination distances are

impossible to measure, because 50 means 50% recombination.

This is a really good point, Sky,

and I appreciate you bringing it up.

50 centimorgan distance means it's just equally probable.

This is 50% recombination.

So it's random chance.

So the reality is that in most of the systems that we use,

the connections really start breaking down

at even much lower, 30 centimorgans, where we're just

seeing a very high rate of double recombination that

is making us systematically underestimate

genetic distances.

And so even yeast recombination breaks down at 50 centimorgans.

So regardless how robust the systems are,

the reality is, in yeast, it's very accurate

up until greater than 40 centimorgans.

And in people or in flies, it's not

accurate at 40 centimorgans.

So this distance-- you are right in that the absolute bar,

the absolute barrier here is 50 centimorgans.

That is correct.

But if we think about a point below there,

say 40 centimorgans it's already broken down

if you look at random gametes, and it's still

a pretty good correlation between physical distance

and genetic distance if you're looking at yeast tetrad

analysis.

Again, I think that the reason why

we talk about this, and the reason

why we introduced this idea is that it's an interesting way

to think about the way that meiosis occurs and introduces

the possibility of what you could see if you could actually

follow all of those recombination events

and see all of the haploids that are generated

from this double recombination event.

And so it's why I think a lot of initial genetics work

and really significant genetics work

has been done in systems like yeast that allow us to study

recombination really well.