/ForkedBiostat-578

A repository for all of my teaching material

Primary LanguageCSS

BIOSTAT 578A: Bioinformatics for Big Omics Data

Instructor: Raphael Gottardo, PhD, Fred Hutchinson Cancer Research Center

If you need to contact me, please email me at rgottard@fhcrc.org.

Time and location: T 10:30-11:50 HST T531 Th 10:30-11:50 HST T747

Prerequisite: BIOSTAT 511/12 or permission of the instructor. Please email me if you're unsure.

Grading scheme (Tentative): HW (40%), Midterm (30%), Final project (30%)

Important dates: Midterm (Feb 20), Final project presentations: last 2 weeks of class (March 4,6,11&13).

Scope: This practical "hands-on" course in Bioinformatics for high dimensional omics will emphasize on how to use statistical methods, as well as the R programming language and the Bioconductor project, as tools to manipulate, visualize and analyze real world omics datasets. The course will be organized around the following topics:

  • Introduction to computing for Bioinformatics using R: Introduction to R/RStudio, review of main data structures and tools for efficient and reproducible research, data manipulation and visualization
  • Managing "big omics data" using relational databases: Overview of main database management systems (MySQL, Postgres, SQLite), and review of the Structured Query Language and main operations
  • How to connect to a database from R, and alternative to databases in R (sqldf and data.table)
  • How to evaluate and adjust the data for presence of "batch effect"
  • Regression techniques for high throughput biomedical data: Multiple regression analysis and logistic regression, ANOVA and design of experiments
  • Statistical methods for high dimensional hypothesis testing: Permutation tests, empirical Bayes and multiple comparison adjustment
  • Modeling of gene expression data: Introduction to Bioconductor, and basic packages for gene expression analysis (GEOquery, Limma, DAVIDquery, etc)
  • Genome-wide association studies and eQTLs; review of main packages in R/Bioconductor (e.g. rqtl)
  • Overview of other high-throughput technologies (e.g. RNA-seq, ChIP-seq) and available tools in R/Bioconductor
  • Data integration: Using R to integrate multiple data types and perform "systems biology" type analysis
  • Drawbacks and limitations of high dimensional omics analysis (overfitting, inference)

Note that this is tentative ouline and minor modifications are likely to occur. Please watch this page regularly for updates.

Lecture notes: