Discovery breast cancer subtypes: unsupervised analysis

M Hallett

21/07/2020

Goals for next two lectures

  • k means clustering (unsupervised analysis) to discover subtypes in breast cancer transcriptomic data.

  • Linear regression models (supervised analysis) to predict patient outcome from breast cancer transcriptomic data.

Breast cancer subtypes

  • Revisit Lecture 3.

From Parker et al. (2010) The PAM50 breast cancer subtypes.

Predicting outcome from transcriptional profiles

  • Revisit Lecture 6 especially the end. There we were classifying tissue type not outcome, but the principle is the same.

BIOL 480

© M Hallett, 2020 Concordia University

Computational Biology Concordia