Overview

Gene expression level is a complex phenotype. Its measurement in an experiment can be affected by a wide range of factors - state of the cell, experimental conditions, variants in the sequence of regulatory regions, etc. To understand genotype to phenotype relationships, we need to be able to distinguish the variation that is due to genetic state from all the confounding causes.

Here, we present PEER, a probabilistic framework for understanding sources of variation in high-dimensional phenotype data. PEER allows models for different sources of variation in the observed phenotype to be quickly and flexibly combined, resulting in more accurate estimates of effect of variable of interest.

Mirror

PEER is also hosted at Sanger
[Sanger]

Download

• Python source code:
  [source]

Relevant Publications

• Oliver Stegle, Anitha Kannan, Richard Durbin and John Winn
  Accounting for Non-genetic Factors Improves the Power of eQTL Studies.
  Lecture Notes in Computer Science (RECOMB) 2008.
  [pdf]
• Oliver Stegle*, Leopold Parts*, Richard Durbin, Anitha Kannan and John Winn.
  A Bayesian framework to account for complex non-genetic factors in gene expression levels greatly increases power in eQTL studies .
  Submitted / working paper 2009.
  [pdf]