Statistical Methods for the Analysis of Microarray and Short-Read Sequencing Data

This lecture is offered in the fall semester 2010 for students of University and ETH Zurich. It is part of the Master in Statistics (ETH) and the Master in Computational Biology & Bioinformatics.

Microarrays and the latest Short-Read Sequencing technologies are the main workhorses to gain insight in the RNA and DNA world of cells and tissues. The main characteristic of both technologies is that they do not only measure single genes or genomic regions but can provide genome-wide measurements in a single experiment. They achieve this by a measurement process that is massively parallelized and can thus interrogate millions of sequences at the same time.

The main application of microarrays is to measure the gene expression or gene activity which is frequently used to identify the changes of the gene activity in cells or tissues induced during development or external stimuli like drug treatments or environmental changes. Many other applications like, e.g. genotyping, do exist but are less frequent.

For Short-Read Sequencing there is not yet a main application, it is equally well suited to

• Measure gene expression
• Identify transcript variants
• Identify genome-wide transcription factor binding sites
• De novo sequencing of new organisms
• Resequencing of organisms
• ...
  

This lecture covers the statistical methods that are used to preprocess and analyze both types of data.

All methods will be exemplified in the exercises using real-world data. The exercises will be conducted using the R programming language. Basic knowledge of the R programming language is required!

The topics of the lecture are

• Data preprocessing: Conversion of raw measurement data
• Exploratory data analysis: Identification of the major data characteristics
• Differential expression: Use hypothesis tests to identify changes in gene expression
• Transcript variation: Identification of alternative usage of the same genomic locus
  
• ChIP-chip or ChIP-seq: Identify genomic regions that are enriched in samples
  
• Classification: Using expression data to build predictive models
• Functional analysis: Mapping genes or genomic regions to biological annotation like functional categories or pathways
  

The lecture is relevant for everybody who has an interest in the areas of applied statistics, bioinformatics or molecular life science.

Course language will be English.

R Courses

This lecture does not teach R itself, but shows how you can apply R to -omics problems if you already know how to write scripts in R. At ETH there are introductory courses to R

• 2009 course given by Prof. W Stahel (with links to R introductory material)
  
• Courses by the ETH Informatikdienste
  

Date & Location

Period: Mon Nov 1 to Mon Dec 13 2010
Lecture: Mondays 09:15 - 11:00, ETH Zentrum, TBA
Exercise: Mondays 11:15 - 12:00, ETH Zentrum, TBA

Entries in the Course Catalogs

ETH: TBA

UZH: TBA

Contact

Hubert Rehrauer