MIT CompBio Lecture 06 - Expression Analysis Clustering Classification (Fall '19)

MIT Computational Biology: Genomes, Networks, Evolution, Health

Prof. Manolis Kellis

Full playlist with all videos in order is here

All slides from Fall 2019 are here

Outline for this lecture:

1. Introduction to gene expression analysis

- Technology: microarrays vs. RNAseq. From reads to transcripts, expr levels

- Expr matrix. Supervised vs. unsupervised learning. Clustering/Classification

2. K-means clustering (clustering by partitioning)

- Algorithmic formulation: Update rule, optimality criterion. Fuzzy k-means.

- Machine learning formulation: Generative models, Expectation Maximization.

3. Hierarchical Clustering (clustering by agglomeration)

- Basic algorithm, Distance measures. Evaluating clustering results

4. Naïve Bayes classification (generative approach to classification)

- Discriminant function: class priors, and class-conditional distributions

- Training and testing, Combine mult features, Classification in practice

5. (optional) Support Vector Machines (discriminative approach)

- SVM formulation, Margin maximization, Finding the support vectors

- Non-linear discrimination, Kernel functions, SVMs in practice