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