Evolution Lab 1, the best Simulation of Genetic Drift in the Universe by Ghaza Hosseini and Kevin Tang :D

In this experiment, we used beans and beads to simulate genetic drift and natural selection. An initial gene pool 100 consisting of 25 round, 25 pinto, 25 black and 25 white bean was created, from which 50 beans were randomly drawn to create the upcoming generation. This process was repeated until 10 generations were obtained and studied. We came up with the null hypothesis that the beans won’t be affected by genetic drift, as well as an alternative hypothesis that the beans will be affected genetic drift. We also thought that natural selection would act on size, favoring the bigger beans.

Figure 1. The data collected from the first set of 10 "generations" from experiment 1.

Beans	Expected	Observed	o – e	(o – e)2	(o – e)2/e
Round	10	36	26	676	67.6
Pinto	11	14	3	9	0.82
Black	13	0	13	169	13
White	16	0	16	256	16
Total	50	50	-	-	97.4

DF = 3                p = 0.05               Critical Value = 7.82

We determined that population A experienced genetic drift as both the black and white beans completely died out in figure 1, while the round beans consistently propagated to a total of 360%, and the pinto bean population rose to 200% but fell back down to almost the original size of 11. We also calculated the chi-square value to be 97.4 for population A, which is far bigger than the critical value of 7.82, so we reject null hypothesis. We also didn’t think this was natural selection because the round beans are the smallest  but came out on top while the pinto beans which are the biggest came in second.

Figure 2. Shows the data collected of the second set of 10 "generations."

Beans	Expected	Observed	o - e	(o – e)2	(o – e)2/e
Round	15	19	4	16	1.07
Pinto	14	19	5	25	1.79
Black	12	6	6	36	3
White	9	6	3	9	1
Total	50	50	-	-	6.86

DF = 3                p = 0.05               Critical Value = 7.82

Genetic drift didn't seem to have occurred in population B. The population of all beans remained relatively steady, with the most significant changes being the black beans' population dropped b 50% and the pinto bean population grew by almost 36%, but no beans died off or significantly increased in population. The chi-square test also supports the null hypothesis that the beans won't be affected by genetic drift because the chi-square value is less than the critical value.

The experiment showed us that genetic drift is completely random, as the two data sets contradicts each other. It's also worth noting that two different techniques were used to create the 10 generations of beans. In population A, beans were randomly drawn from a cup, so size will influence the chances of being picked. In population B, beans were mixed, then dumped out of a cup and the pile was cut in half to be counted and used. Perhaps in future experiments, there should be more than two sets of data and sorting techniques should be controlled to maximize accuracy.