Chi-square - Determining Probability Values from Genotypic Data

Deana Namuth-Covert Author

01/13/2019 Added

7400 Plays

Description

This video clip demonstrates how to calculate the expected F2 frequencies and numbers of individuals based upon genotypic data from DNA markers.

Searchable Transcript

Search:

[00:00:00.160]Hello, this is Dr. Deana.
[00:00:09.640]In this video clip, we'll be continuing on with our chi-square calculation using genotypic
[00:00:15.960]data.
[00:00:17.600]You'll see here I've copied the information from the table that's in your lesson that
[00:00:22.340]you're working through, and you'll see that we've got our three different classes.
[00:00:28.120]We've got our F2 individuals, which had a banding pattern similar to OH88119.
[00:00:36.360]We also have our class in which the banding pattern was similar to our other parent, 6.8068.
[00:00:44.000]And I've left blank some of the numbers here for those individuals which had the heterozygous
[00:00:49.720]banding pattern, meaning they had all three bands in the gel.
[00:00:55.500]Now in the last video clip, we talked about how to calculate
[00:00:58.000]the expected numbers of F2 plants for each different banding
[00:01:04.040]pattern.
[00:01:05.040]So we'll go ahead and continue on from there.
[00:01:07.760]Okay so what we want to do is calculate our deviation, which is simply the observed number,
[00:01:15.440]so the number of individuals that we saw in our gel which had the banding pattern similar
[00:01:22.180]to the heterozygous condition that we were expecting, so that was 66, and then we're
[00:01:27.880]going to subtract out the number we expected, which is 98.5, and when you punch that out
[00:01:35.360]you're going to come up with a negative 32.5, so that's our deviation.
[00:01:44.800]Then our next column, what we're going to do is simply square this number, so minus
[00:01:49.060]32.5 times minus 32.5, and that will give us a positive 1,056.25.
[00:01:57.760]So again, that came from negative 32.5 times negative 32.5.
[00:02:13.420]Then in our last column, we're going to take this number, our observed minus expected squared,
[00:02:19.120]and divide it by our expected, and we're going to divide it by the expected number
[00:02:24.760]for this particular class.
[00:02:27.640]We're going to be in the same row, so 1056.25 divided by 98.5 gives us 10.7234.
[00:02:41.800]Again, that was 1056.25 divided by 98.5.
[00:02:53.720]We need to then add up all three chi-square values
[00:02:57.520]in order to get our total chi-square, so we'll add all of those, and when you do that you
[00:03:06.140]end up with 36.8884.
[00:03:18.240]Our total chi-square value then for this experiment would be 36.8884.
[00:03:27.400]Let's look at this number then and look it up in our probability chart, but before we
[00:03:31.980]do that, remember we also need to know what our degrees of freedom are.
[00:03:39.760]We need to know what our degrees of freedom number is, and you'll recall that that is
[00:03:46.960]the number of classes in your study minus one.
[00:03:57.280]So how many classes do we have here? Well, we have three.
[00:03:59.780]Because we've got genotypes that give us a banding pattern similar to the OH parent.
[00:04:07.980]We have genotypes that give us a banding pattern similar to the 6.8068 parent.
[00:04:14.400]And then we also have some that are showing the heterozygous condition,
[00:04:18.760]which is a combination of banding patterns.
[00:04:21.780]So three classes minus one gives us two degrees of freedom.
[00:04:28.120]All right, so remember the two degrees of freedom and our 36.8884.
[00:04:36.320]And let's go ahead and look then at our probability distribution table.
[00:04:40.500]Okay, so remember what we're gonna do is read down this [degrees of freedom] column
[00:04:45.780]until we hit the two degrees of freedom.
[00:04:48.360]Let's go ahead and zoom in a little bit more.
[00:04:51.840]Okay, so we have two degrees of freedom [as our row].
[00:04:55.780]So we're gonna read across in this one.
[00:04:58.100]So we're gonna read across this row until we find 36.884.
[00:05:05.100]Holy cow, this is like clear off the chart.
[00:05:08.180]Okay, so we're gonna keep going to the end and
[00:05:11.180]it's gonna be far, far less than 0.01.
[00:05:14.220]So does this mean we can accept our hypothesis or fail to reject it?
[00:05:20.880]Or should we reject our hypothesis?
[00:05:22.780]Okay, well if you said reject, then you are correct.
[00:05:27.480]So we're gonna reject our hypothesis.
[00:05:32.020]And that's because our cutoff value is this 0.05.
[00:05:38.840]Okay, remember anything to the right of this [0.05 value], and
[00:05:43.100]we're gonna have to reject our hypothesis.
[00:05:45.280]Okay, here I've got a little bit cleaner drawn out [dotted line for 2 degrees of freedom appears].
[00:05:49.460]All right, so remember our hypothesis was that our molecular marker COS
[00:05:57.200]OH57 was co-dominantly inherited in our tomato population.
[00:06:07.240]And in that tomato population, we had our original parents of
[00:06:12.820]OH88119 times the 6.8068 parent.
[00:06:18.620]Okay, so this particular marker in a population from these two
[00:06:27.180]parents, we were expecting a 1:2:1 ratio.
[00:06:35.260]Or one-fourth of the banding patterns would be like the OH parent.
[00:06:39.940]One-fourth would be like the other parent.
[00:06:42.020]And then a half of all the individuals would be a combination of the two
[00:06:46.800]with the banding pattern.
[00:06:48.620]But we have to reject our hypothesis, so this is not correct.
[00:06:53.120]And there's probably some other kind of genetic factors that are
[00:06:57.160]involved, so some other genetic principle or forces is at play.
[00:07:03.280]Okay, so it could be that this marker is actually linked to our disease
[00:07:11.940]resistance trait, or it could be that we have something like natural selection
[00:07:17.500]that has gone on to where this particular marker is,
[00:07:22.680]the banding patterns are at different frequencies than what we would expect.
[00:07:26.160]Bottom line is you have to
[00:07:27.140]go back to the drawing board and revisit your hypothesis and test a new one.
[00:07:32.380]Okay, so this wraps up then our chi-square tutorials.
[00:07:39.760]And so this one was showing you how you actually calculate a chi-square value
[00:07:45.260]based on genotypic data, and then how you interpret that chi-square value.
[00:07:49.780][music]

The screen size you are trying to search captions on is too small!

You can always jump over to MediaHub and check it out there.

Comments

0 Comments

Chi-square - Determining Probability Values from Genotypic Data

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments