Your name: Your email address:
Examine the two sets of raw intensity files Noggin9 and Noggin20. Check MA-plots and histogram/density plots using Limma and ArrayQuality package from BioConductor/R.
Question 1: Determine which slide(s) of the two sets has poorer quality and require further preprocessing before statistical analysis?
For your preferred dataset of Ex.1, do the following three steps and then Q/C: a. Background correction alone and Q/C b. Normalization alone and Q/C c. Background correction followed by normalization and Q/C
Q2: Do you find difference between the 3 preprocessing procedures? What is the effect of background correction and/or normalization on the dataset? Which method among a, b and c should be used on your dataset and why?
For your preferred dataset of Ex.1, use each of the following three methods to determine differentially expressed (DE) genes: a. SAM b. Mannova c. SPH
Q3: Enter the number of DE genes picked by each method.
For your preferred dataset of Ex.1, use each of the following method to conduct clustering: a. hierarchical clustering and consider what size of k you would use b. k-means with the k value of a.(run 3 times)
Q4: What was your choice of k? In b, do you find your three results all consistent? Why do you think they are different?
Download the DE gene list from here, check against KEGG and examine up-regulated and down-regulated genes in the pathways. KEGG tool is availalbe here. Click on Launch Kegg Search to start the program.
Q5: Is wnt pathway affected in this experiment? What other pathway(s) do you think is activated or deactivated in the experiment?
Consider the new data set FGF. Repeat the steps of Ex 1 - Ex 3 and determine whether applying either or both background correction and normalization makes the DE gene selection different. Do you agree that preprocessing is a necessity?
Q6: Please paste your regulated gene list. If time permits, enter the regulated gene list from all three methods, SAM, Mannova and SPH.
Send email to your instructor (and yourself) upon submit Send email to yourself only upon submit (as a backup) Show summary upon submit but do not send email to anyone.