Cancer Genomics Module

GENOME 541 A Sp 23 Course Website

The Cancer Genomics Module is the third module in the course and consists of 4 lectures.
Homework #7 and #8 accompany this module.

Schedule of lectures and homework

Dates: May 9 - May 18

Times: Tuesday & Thursday @ 10:30am - 11:50am

Location: Foege S110

Instructor: Gavin Ha (gha@fredhutch.org)

Homework

Module Outline

Lecture 1: Introduction to cancer genome analysis

Lecture 1 Slides

1. Background on Cancer Genome Alterations
   • Genomic alterations in cancer: drivers vs passengers, somatic vs germline
   • Tumor evolution and heterogeneity
2. Overview of Cancer Genome Analysis
   • Computational strategy and workflow
   • Tumor DNA Sequencing
   • Types of genomic alterations predicted from tumor sequencing
   • Methods/tools/algorithms in following lectures
3. Primer on statistical modeling
   • Probability distribution, Bayesian statistics, inference

Lecture 2: Probabilistic methods for mutation detection

Lecture 2 Slides

1. Primer on statistical modeling (cont’d)
   • Mixture models and inference using the EM algorithm
2. Detecting Mutations in Cancer Genomes
   • Visualizing somatic vs germline SNVs
   • Sequencing read count data
   • SNV genotyping strategy
3. Mixture Models for SNV Detection
   • SNVMix probabilistic model and EM inference
   • Predicting somatic SNVs in cancer

• References:
  SNVMix: Goya et al. Bioinformatics (2010)
  JointSNVMix: Roth et al. Bioinformatics (2012)

Lecture 3: Probabilistic methods for copy number alteration detection

Lecture 3 Slides

1. Detecting Copy Number Alterations in Cancer Genomes
   • Predicting copy number features from sequence data
   • Copy number analysis workflow
   • Data normalization
2. Continuous Hidden Markov Model (HMM)
   • Graphical model representation
   • Components of a continuous HMM
   • Inference & parameter estimation using expectation-maximization (EM)
3. Copy Number Profiling using a Hidden Markov Model
   • Probabilistic model for copy number analysis
   • Predicting copy number segments using the Viterbi algorithm

• References:
  HMMcopy: Ha et al. Genome Res (2012)
  ichorCNA: Adalsteisson et al. Nat Commun (2017)

Lecture 4: Additional topics

Lecture 4 Slides

1. Additional Copy Number Analysis Features
   • Allelic copy number analysis
2. Estimating tumor heterogeneity
   • Modeling tumor-normal admixture
   • Modeling tumor clonality and heterogeneity
3. Assessing Statistical Power for Variant Discovery
   • Power analysis
   • Calibrating sequencing depth for variant discovery
4. Structural Rearrangement Analysis in Cancer Genomes
   • Structural variant types predicted from sequencing analysis
   • Complex genomic structural rearrangements

• References:
  ichorCNA: Adalsteisson et al. Nat Commun (2017)
  TitanCNA: Ha et al. Genome Res (2014)