Statistics for Systems Biology 2010

Description

The course introduces statistical methods and underlying concepts for data analysis with a focus on systems biology. Topics covered in the course include parameter estimation, hypothesis testing, multiple testing problems, and experimental design. The course emphasizes modern computational approaches using the statistics software R.

Objective

The aim of this course is two-fold: first, students should be introduced to standard methods from statistics with application to systems biology datasets of medium complexity. For applying these methods, students will use the software R for assignments. Second, students should gain a good understanding of the underlying principles and concepts in order to be able to choose from the vast set of available methods and critically employ them. In particular, the course will try to avoid tedious computations in favor of a general understanding how methods work and when they will fail. Additionally, the role of modeling a statistical problem and deciding upon a strategy for analyzing a dataset will be favored over presenting as large a number of tests and estimators as possible.

Content

(1-2) Concepts from probability theory: events, probability, random variables, distributions, moments, stochastic independence, joint probabilities; (3-6) Parameter estimation: estimation problem, maximum likelihood estimators, Bayesian estimators, comparing estimators; (7-10) Hypothesis testing: test problems, type-I and type-II errors, power, p-values, multiple testing; (11-12) Statistical design of experiments; (13) Summary and open problems.

Lecture Material

The protected pagelecture materiallock is available here.