Biology Labs

Allele Shuffling of Single Gene Traits in Drosophila

Jasper Graves

Biolab

2/8/12

Introduction:

            This lab is being done to determine whether single gene traits are random or assigned because of some other factor. The experiment was conducted with fruit flies because their small size and ease of storage, and they are bred to the point where you can order flies with isolated and individual traits. These flies were used because they have been studied to great extent and there is a great deal of information to answer most questions. The null hypothesis for this experiment is: If the p-value is greater than 5% than allele shuffling can be said to be random.

Materials:

            The experiment subjects were common fruit flies. The flies lived and bred in plastic test tubes. When the flies are to be counted for traits and gender, they were paralyzed with fly nap(<1oz). The flies were fed with potato mash. Bacteria in the test tubes were killed with baking yeast. Microscopes were used to sort flies with less obvious traits. Flies were moved on petri dishes with small paint brushes. Alcohol was used to dispose of flies (1oz).

Methods:

            First the class studied Flynaped samples of wild-type and mutant flies to determine what the  different traits were and which ones were dominant (microscopes were used when necessary). These flies were than killed using alcohol. Once the offspring (F1) of the first cross were ready to be examined they were Flynaped and taken from their test tubes and sorted using microscopes and small paintbrushes. Than a test tube was prepared for breeding the three different crosses, this was done by filling the tube partially with potato mash and five grains of yeast. Than five unconscious male flies and five unconscious female flies were placed in the tube and sealed. The remaining flies were then placed into the alcohol. Punnet Squares were conducted for the F1 cross. After larva had been laid in the potato mash the parents were removed and placed in the alcohol. After all of the F2 generation was hatched they were anesthetized with Flynap and removed from there test tubes to be sorted for the last time. Totals for the different traits were calculated and Punnet Squares were drawn to determine the ratios for how many flies would have which traits. All flies were then placed in the alcohol.

Results and Analysis:

Sample calculation for expected number of flies: Percent of flies with given trait X Total number of flies with trait.

Sample calculation for Chi-Square value: Total of (observed – expected) to the second power divided by expected.

All of the crosses were random except for vestigial red and vestigial sepia from the VvSs x VvSs cross.

Ss x Ss
Phenotype	Observed	Expected	Chi-Squaure value	p-value	Significant
Red-eyed	326	299.25	2.391	>5%	N
Sepia-eyed	871	897.75	0.797	>5%	N
Total	1197
VvSs x VvSs
Phenotype	Observed	Expected	Chi-Squaure value	p-value	Significant
vestigial red	197	258	14.422	<5%	Y
vestigial sepia	231	86	244.477	<5%	Y
normal red	708	774	5.628	>5%	N
normal sepia	240	258	1.256	>5%	N
total	1376
Xw+Xw- x Xw-Y
Phenotype	Observed	Expected	Chi-Squaure value	p-value	Significant
red female	214	217.5	0.056	>5%	N
red male	240	217.5	2.333	>5%	N
white female	184	217.5	5.159	>5%	N
white male	228	217.5	0.507	>5%	N
total	866

 

Conclusion: If the p-value is greater than 5% than allele shuffling can be said to be random. If the data showed that all the fly traits were random, than all the data would have come up >5% (Chart above). Our data showed instead that the phenotypes vestigial red and vestigial sepia were not random and were somehow effected by an out side source. Our hypothesis is not correct, not all of the traits are random. This may be do to several reasons, our flies were not kept at a constant temperature and some of the data collected when counting the traits may not have been accurate. With flies this is not that significant because having different eye color for instance does not change how the flies live; but for humans who want a prediction for what the traits of there children will be this could be damaging. For instance, if one of the parents has hemophilia and wants to know the likelihood of passing it on to their child gets a prediction that there is is only a 25% chance of it happening; but if predictions do not work than this means the parent really dos not know what the chance of their child getting hemophilia is. I feel that some of the statistics I have may not be completely accurate, human error is to blame I think that some of the counting was done with out precision. This experiment should be followed up with another one that will try and explain why the flies did not follow their punnet square results.