Time to try it at home!

Here a few exercises based on the example presented in the lesson:

  1. Try to replicate the analysis using the brainarray CDF. Do you arrive to the same conclusions?

The following exercises can be done either with the original data or with the brainarray data

  1. t-SNE is a relatively new dimensionality reduction method that allow to visualize high-dimensional data in 2D. What type of conclusion we get if we use t-SNE instead of PCA? Note that t-SNE is implemented in the package tsne, and is described in this publication:

L.J.P. van der Maaten and G.E. Hinton. Visualizing High-Dimensional Data Using t-SNE. Journal of Machine Learning Research 9(Nov):2579-2605, 2008

  1. Try to use classical linear modelling, i.e., the lm() function instead of limma. Do the results change?

  2. Do not scale the data before plotting the heatmap. How do the plot change?