Matrix Decomposition for Data Integration lab comes in two flavors, light and detailed version. See a short description on what to except from each and feel free to choose the one you want to work with during the workshop.

Light version

  • Shows how we use mixOmics package for unsupervised dimensionality reduction, PCA (one data set) and GCCA (multiple data sets).
  • Shows how we use mixOmics package DIABLO method for supervised data integration.
  • Gives minimalist explanation of methods, skipping mathematical foundations. Maybe be good as first read or as a quick guide how to use mixOmics when already knowing the foundations behind the methods.

Take me to the light version

Detailed version

  • Introduces unsupervised GCCA method starting from basics by successively building upon PCA method and CCA. Gives mathematical foundations, R code from scratch as well as demonstrates how things can be run via mixOmics package.
  • Builds further upon GCCA to introduce PLS and PLS-DA and their explanations to supervised integration. R code from scratch and via mixOmics.
  • May take some time to go through and be slightly challenging. Recommended when wanting to understand the methods.

Take me to the detailed version