How many dyes are usually used?
Given the usual choices, what does yellow correspond to in the generated images?
Why is a two-fold increase in expression not a good measure of significance? (You could answer this question giving an example)
If you want to compare more than two different conditions, how could you label the different samples? Which would you use in the same hybridization experiment? Describe at least 3 different experimental designs. If you use a diagram with arrows (which would be an excellent way to address this question), explain what the arrows stand for.
How might one use DNA Microarrays to find genes that are involved in the same metabolic pathway?
What might go wrong using this approach (think false negatives)?
Describe how you would proceed using a DNA microarray to search for shared regulatory elements that might be involved in the response of human tissues to lack of oxygen.
The ratios (Ti) of red (Ri) to green (Gi) fluorescence intensity usually are given as logarithms to the base of 2.
If the red fluorescence is 4 time more intense than the green fluorescence, what is the log(base2)Ti?
If the green fluorescence is 4 time more intense than the red fluorescence, what is the log(base2)Ti?
If the red fluorescence is 8 time more intense than the green fluorescence, what is the log(base2)Ti?
If the red fluorescence has exactly the same intensity as the green fluorescence, what is the log(base2)Ti?
Why might the DNA arrays manufactured by Affymetrix might yield more reproducible results than the DNA arrays printed onto surface treated glass slides?
To measure reproducibility, one often performs replicates. These can be either technical (the same probes are used on different chips, or the same cDNA is used in two separate spots), or biological. Why might one type of replicate be preferable over the other.
Imagine that you analyze the expression of 1000 mRNAs under two conditions. Assume that you perform sufficient replicates to assess the variance for each mRNA. A simple ANOVA test of the samples for a single mRNA allows you to decide if the difference between the two conditions is significant. If you were to apply this test to all 1000 mRNAs, how many false positives would you expect, if you were to perform your test at a significance level of 5%?
If one were to apply the Bonferroni correction for multiple tests, at what significance level would one need to perform the individual tests? (Aside: the null hypothesis tested at the 5% significance level would be: None of the mRNAs shows significant change.)