Foundational Computational Biology I-II (2022)

Graduate course, Department of Molecular Genetics, University of Toronto. Note: only lectures given by me are listed below.

Lectures

• [1.1.2] An introduction to reproducible computational research [slides]
  • Why work reproducibly?
  • Pseudorandom number generation and seeds
  • An introduction to Snakemake
• [1.3.2] An introduction to supervised machine learning [slides]
  • An overview of supervised learning
  • Linear regression models
  • Loss optimization via gradient descent
  • Classification with logistic regression
  • More complex models
  • Train/test splits
  • Model complexity: overfitting and underfitting
  • Penalized regression
• [2.1.1] Unsupervised learning: continuous latent variable models [slides]
  • Historical perspective
  • Principal component analysis and pPCA
  • Non-negative matrix factorization
  • Nonlinear methods: tSNE & autoencoders
• [2.1.2] Unsupervised learning: mixture models [slides]
  • Clustering from a probabilistic perspective
  • Gaussian Mixture Models
  • Expectation-maximization
  • Model selection
• [2.3.1] Introduction to Bayesian inference [slides]
  • Re-introduction to Bayes rule
  • Sampling methods: Gibbs sampling, Metropolis Hastings
  • Variational inference
  • An introduction to probabilistic programming languages & STAN
• [2.3.2] Introduction to deep learning [slides]
  • The perceptron, multi-layer perceptrons
  • Gradient descent and backpropagation
  • Deep learning for images: CNNs
  • Deep learning for sequence data: RNNs