R Methy V: Normalization

The Minfi package has several possible normalization functions, depending on which method you prefer.

• Perform a normalization, using one of the available methods. One is more suitable than another based on the dataset, use the MDS plot to support your choice. Put your normalized data in a variable named minfi.norm. If your chosen normalization method does not perform background correction, first perform BC using preprocessNoob(). Put your normalized data in a variable named minfi.norm.
• If your chosen normalization function also performs a background correction, which BC method is used?

Note:

Because of a bug in minfi, if you use preprocessNoob followed by preprocessQuantile you need to add the following:

myMinfiBC@preprocessMethod <- c(rg.norm = sprintf("Noob, dyeCorr=%s", TRUE))

where myMinfiBC contains the background corrected data.

Questions:

6. Minfi also has a function to check the gender of each sample, plotSex(). Apply this function to your QC object and compare the plot to your sample sheet. Hand in the plot and your conclusion of the comparison.
7. Which normalization method did you use with minfi?

---

Next: Analysis

Previous: Quality Control