From OpenWetWare
Jump to: navigation, search

Statistics for Microarrays: How to make good inference

Author(s): Ernst Wit
Affiliations: The University of Lancaster
Keywords: microarray; statistics; tutorial


Microarray experiments have become a model for how new breakthroughs in high-throughput biotechnologies will call upon computational scientists for help. Data from such experiments are high-dimensional, noisy, but form at the same time a highly structured system, about which other related information (sequence data, gene ontology, proteomic data, metabolic data) is available. Statistically inclined scientists are the type of computational scientist ideally suited to deal with these types of experiment.


In this tutorial, we will touch upon a range of inferential aspects for which experimentalists typically need the help:

  • What does it mean to find differentially expressed genes?
  • How to control the False Discovery Rate?
  • How to find genes that predict class membership?
  • Can microarray data play a role in modelling molecular biological processes?

Workshop Prerequisites

Could all attendees ensure that they come with the following:

  • Laptop computer
  • R installed on the laptop
 R can be downloaded from the following location:
  • Download the following Data Files into a NEW directory
  • Download the following PDF file onto their laptop: