Check your mixed vial, and if it has a decent number of adults, transfer them to a food vial. If you tranferred adults last Thursday, remove them and count them.
Work on you individual project. If you have data, talk to me about how to analyze it statistically.
Calculate the allele frequency for your mutant allele at each generation you have data for. For autosomal genes, calculate the allele frequency in the total population, using the Hardy-Weinberg relationship. For sex-linked genes, calculate the allele frequency separately for each sex, then take the average of the two numbers.
It is unlikely that the allele frequency is exactly the same for the first generation and for your more recent generation. There are two possible reasons for a change in allele frequency: random drift, or natural selection. To see whether random drift could have caused the change you see, use this spreadsheet to simulate random drift.
Enter your allele frequencies for each generation that you have data for, plus the number of flies you counted at each generation. Then hit the "control" and "equal" keys. This will make the spreadsheet simulate random drift, with the same starting allele frequencies as in your data, and the same number of flies at each generation.
The graph on the spreadsheet will show the simulated change in allele frequencies in blue. If it ends up being as far as, or farther from, the starting allele frequency as your observed data were, the spreadsheet will say "MORE EXTREME."
Run the simulations 100 times, by hitting the "control" and "equals" keys 100 times. Each time you hit these keys, the spreadsheet will run a new simulation. Count how many times, out of the 100, that the simulated data show a difference between the start and end that is as big as the one you observed. If you get a lot (more than 5 percent), it suggests that random drift could be responsible for the change in allele frequency that you see.
If fewer than 5 percent of the simulated changes in allele frequency were as large as the observed one, change the "Ne/N ratio" to 0.5. This means that in the simulations, only one-half of the flies will actually reproduce. Then do 100 more simulations. The effective population size (Ne) is often quite a bit smaller than the census population size (N); this will give you an idea of how small the population would have to be for drift to produce the kind of difference you've seen.
If more than 5 percent of the simulated changes in allele frequency were as large as the observed one, change the "Ne/N ratio" to 2. This will see whether random drift could have produced the kind of change in allele frequency you've seen, even if the population were twice as big.
To simulate selection, use this spreadsheet for selection on autosomal genes or this spreadsheet for selection on sex-linked genes. Try different values of the relative fitness until you get a difference as big as the one you've seen.
Return to the Genetics Lab syllabus
Return to John McDonald's home page
This page was last revised October 27, 2010. Its URL is http://udel.edu/~mcdonald/geneticslab17.html