GPrank: Gaussian Process Ranking of Multiple Time Series

Implements a Gaussian process (GP)-based ranking method which can be used to rank multiple time series according to their temporal activity levels. An example is the case when expression levels of all genes are measured over a time course and the main concern is to identify the most active genes, i.e. genes which show significant non-random variation in their expression levels. This is achieved by computing Bayes factors for each time series by comparing the marginal likelihoods under time-dependent and time-independent GP models. Additional variance information from pre-processing of the observations is incorporated into the GP models, which makes the ranking more robust against model overfitting. The package supports exporting the results to 'tigreBrowser' for visualisation, filtering or ranking.

Version: 0.1.4
Depends: R (≥ 2.14.0)
Imports: gptk, matrixStats, tigreBrowserWriter, RColorBrewer
Suggests: knitr, rmarkdown
Published: 2018-08-17
Author: Hande Topa [aut, cre], Antti Honkela [aut]
Maintainer: Hande Topa <hande.topa at>
License: MIT + file LICENSE
NeedsCompilation: no
Citation: GPrank citation info
CRAN checks: GPrank results


Reference manual: GPrank.pdf
Vignettes: GPrank vignette
Package source: GPrank_0.1.4.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X binaries: r-release: GPrank_0.1.4.tgz, r-oldrel: GPrank_0.1.4.tgz
Old sources: GPrank archive


Please use the canonical form to link to this page.