Repository navigation
R Expr III: Quality control
Your data is currently saved as an AffyBatch object, which cannot be handled by limma. The function rma() in the package affy can do background correction, normalization (quantile normalization) and return an ExpressionSet object which limma can take care of. You can also decide to skip the background correction and/or the normalization using the normalize and background arguments, and you will still get an ExpressionSet” object.
- Apply the
rma()function to your raw data both with and without normalization and background correction, and save it as two new variables with suitable names.
You can access the expression values of your ExpressionSet object using the exprs() function. The boxplot() function does not take an ExpressionSet object as input since it expects a data frame or a matrix, but if you apply the exprs() function to your ExpressionSet it works fine.
2. Use the boxplot() function together with the exprs()function to create boxplots on your each of your two ExpressionSet objects. Also use the main argument to set suitable titles to the two plots. You can also try to do the same with data that is normalized but not background corrected, or the other way around. At the top of the plot window in R-Studio there is a button called Export that you can use to save your plots. The boxplots of both your data sets (with and without background correction and normalization) should be included in your report.
3. Use the plotMDS()to create two MDS plots on the same data as for the boxplots. The plotMDS() function can read ExpressionSet objects so you don't need to use the exprs() function here. The two MDSplots should be included in your report.
4. Write a comment in your report about the differences you can see before and after normalization and background correction.