Skip to content
ostegle edited this page Oct 31, 2011 · 33 revisions

##What is PEER## PEER is a collection of Bayesian approaches to infer hidden determinants and their effects from gene expression profiles using factor analysis methods. Applications of PEER have

  • detected batch effects and experimental confounders
  • increased the number of expression QTL findings by threefold
  • allowed inference of intermediate cellular traits, such as transcription factor or pathway activations

The PEER model, inference, and applications are described in

A protocol paper, describing the use of PEER and this software package is described in

  • Using Probabilistic Estimation of Expression Residuals (PEER) to obtain increased power and interpretability of gene expression analyses (Nature Protocols, int press)

and several other projects have successfully used PEER.

This project offers an efficient and versatile C++ implementation of the underlying algorithms with user-friendly interfaces to R and python. To get started using PEER, download the source or binary versions, see the installation instructions, and take a look at the getting started tutorial.

##News##

  • 10/10/2011 PEER 1.3 release: improved verbose functions for the R interface and new demo scripts.
  • 06/08/2011 PEER 1.2 release: minor bugfixes and interface additions for R and python interface.
  • 01/08/2011 Extended examples online, showing how to use PEER on yeast eQTL datasets
  • 20/07/2011 PEER 1.1 release: PEER can now also be installed as R source package (CRAN forthcoming)
  • 14/07/2011 PEER 1.0 released: Support for sparse factor analysis with prior information added
  • 01/07/2011 Support for probe-specific measurement uncertainty added

##Links##

##Who is behind PEER##

  • Oliver Stegle
  • Matias Piipari
  • Leopold Parts

PEER was originally developed in research groups of John Winn at Microsoft Research, Cambridge, and Richard Durbin at the Wellcome Trust Sanger Institute.

Clone this wiki locally