Lab 1: Population Genetics with Beans & Beads by Julia Smith, Nia Parvini, and Na Nguyen
Part A: Distribution of Genes in a Population
Hypothesis: When selected and paired together at random, sample alleles will produce a sequence
of genotypes consistent with mendelian ratios.
of genotypes consistent with mendelian ratios.
Prediction: Of the 50 bean pairs, 25% will be homozygous white, 25% will be homozygous speckled,
and 50% will be heterozygous.
and 50% will be heterozygous.
Data:
Figure 1. Genotype frequencies observed in the random pairing of model alleles (white beans and speckled beans).
Analysis: As our group’s chi-squared value of 10.24 exceeds the critical value of 5.99,
we must reject our null hypothesis and consider which factors other than chance may have played
a role in our data collection. Rather than the expected values of 12.5 homozygous white pairs,
12.5 homozygous speckled pairs, and 25 heterozygous pairs—calculated according to mendelian ratios—
our observed values—shown in Figure 1—in both trials were each closer to ⅓ of the total pairs.
Due to the absence of alternative explanations, the most likely reason for the discrepancy
between our expected and observed results is human error in the randomization process
used to select the bean pairs. We failed to account for the perceptible difference in size and shape
between the white and speckled beans, so by selecting the beans by hand we may have allowed
unconscious biases towards one type of bean or the other to skew the pairing ratios towards
homozygous genotypes and away from the heterozygous genotype. Compiling the data from the
entire lab section produced a chi-squared value of 12.876, which is also greater than 5.99,
indicating that other groups in the class likely made similar mistakes. It may be beneficial for our group
to repeat this experiment using a selection method which eliminates the sense of touch as a factor in
order to see if our data become closer to mendelian ratios.
we must reject our null hypothesis and consider which factors other than chance may have played
a role in our data collection. Rather than the expected values of 12.5 homozygous white pairs,
12.5 homozygous speckled pairs, and 25 heterozygous pairs—calculated according to mendelian ratios—
our observed values—shown in Figure 1—in both trials were each closer to ⅓ of the total pairs.
Due to the absence of alternative explanations, the most likely reason for the discrepancy
between our expected and observed results is human error in the randomization process
used to select the bean pairs. We failed to account for the perceptible difference in size and shape
between the white and speckled beans, so by selecting the beans by hand we may have allowed
unconscious biases towards one type of bean or the other to skew the pairing ratios towards
homozygous genotypes and away from the heterozygous genotype. Compiling the data from the
entire lab section produced a chi-squared value of 12.876, which is also greater than 5.99,
indicating that other groups in the class likely made similar mistakes. It may be beneficial for our group
to repeat this experiment using a selection method which eliminates the sense of touch as a factor in
order to see if our data become closer to mendelian ratios.
Part B: Genetic Drift or Natural Selection
Hypothesis: The null hypothesis is that the alleles will not be affected by genetic drift and will stay
constant from G1 to G10. The alternative hypothesis is that genetic drift will change the frequency
of alleles between G1 and G10.
constant from G1 to G10. The alternative hypothesis is that genetic drift will change the frequency
of alleles between G1 and G10.
Prediction: Random variations in frequency will occur to the model alleles in G10 compared to the
frequencies observed in G1.
frequencies observed in G1.
Data:
 |
Figure 2. Alleles in G2 of population A.
|
 |
Figure 3. Alleles in G10 of population A.
|
Table 1. Numbers of model alleles (red beads, white beads, clear beads, and black beans) obtained after randomly dividing each population in half and doubling each allele to return it to its original size. Population A originally used 50 total alleles for each generation but all data from this population have been doubled to match the 100 total alleles used in population B.
Figure 4. The effect of random selection of alleles on relative allele frequencies in populations A and B over 10 generations.
Analysis: From the data shown in Figure 4, it would appear that both of the populations were affected
by genetic drift. Looking at it from a statistics angle, we calculated the chi-squared for both of the
population. Population A was 45.12 and Population B was 61.05. Since both of these values exceed
the critical value of 7.82, we can conclude that the null hypothesis can be rejected and that the
alternative hypothesis has been supported by these data. The variations in allele frequency
observed over these 10 generations in both populations are most likely due to genetic drift rather
than natural selection. Natural selection would require that some inherent characteristics of these
alleles made them more or less likely to be selected in each generation, and as we divided each
population without looking at or touching the beads and beans, it is highly unlikely that this was the case.
Rather, it was simply random chance that caused the black beans to be lost in population B and
nearly lost in population A, while the red beads drifted towards fixation in population A and the clear
beads did the same in population B. These results are consistent with the effects of genetic drift.
by genetic drift. Looking at it from a statistics angle, we calculated the chi-squared for both of the
population. Population A was 45.12 and Population B was 61.05. Since both of these values exceed
the critical value of 7.82, we can conclude that the null hypothesis can be rejected and that the
alternative hypothesis has been supported by these data. The variations in allele frequency
observed over these 10 generations in both populations are most likely due to genetic drift rather
than natural selection. Natural selection would require that some inherent characteristics of these
alleles made them more or less likely to be selected in each generation, and as we divided each
population without looking at or touching the beads and beans, it is highly unlikely that this was the case.
Rather, it was simply random chance that caused the black beans to be lost in population B and
nearly lost in population A, while the red beads drifted towards fixation in population A and the clear
beads did the same in population B. These results are consistent with the effects of genetic drift.
This is very interesting. My group were assigned the beans and we have came to the conclusion that the demonstrated evolutionary process was natural selection due to the variation in the sizes of beans in the second exercise . However, I was always curious about the outcome results for the beads in the second exercise. I was expecting it to reject the null hypothesis and demonstrate genetic drift, but the population B support the null hypothesis. This outcome could be due biases in selection and the beads were not randomly selected. For the first exercise, my result and your result followed the mendelian ratio, and supported the null hypothesis. And I can't think of any reasons of why we couldn't prove other wise. It could be due to the population size and lack of variability in the gene pool.
ReplyDeleteReally interesting data for part B! To have the different beads finish around the same point after being so skewed is very unlikely and interesting, especially when you consider the fact that most groups had, or almost had, certain bead colors almost go extinct.
ReplyDeleteLooking at your groups data from part B is fairly interesting due to the fact that population A was rejected by the null hypothesis and population B was accepted. In population A, you can see that the allele frequencies are fluctuating every generation and not being close to the expected data at all. On the other hand, population B has stayed to close to the expected allele frequency for most of the generations and has the same or close to the same allele frequency as the first and last interval. I am truly shocked by the genetic drift results.
ReplyDeleteWe ended up with a similar drift as in your Figure 1, where one of the populations went towards being fixed and another went towards being lost. It is interesting in your figure 2 representation, where your populations drifted and then later converged again. This is a good example where the chi squared test would not provide sufficient evidence in rejecting the null hypothesis.
ReplyDelete